aboutsummaryrefslogtreecommitdiff
path: root/drivers/net/ethernet/intel
AgeCommit message (Collapse)AuthorFilesLines
2021-10-11ice: cleanup rules infoVictor Raj2-10/+59
Change ICE_SW_LKUP_LAST to ICE_MAX_NUM_RECIPES as for now there also can be recipes other than the default. Free all structures created for advanced recipes in cleanup function. Write a function to clean allocated structures on advanced rule info. Signed-off-by: Victor Raj <[email protected]> Signed-off-by: Wojciech Drewek <[email protected]> Tested-by: Sandeep Penigalapati <[email protected]> Signed-off-by: Tony Nguyen <[email protected]>
2021-10-11ice: allow deleting advanced rulesShivanshu Shukla2-0/+229
To remove advanced rule the same protocols list like in adding should be send to function. Based on this information list of advanced rules is searched to find the correct rule id. Remove advanced rule if it forwards to only one VSI. If it forwards to list of VSI remove only input VSI from this list. Introduce function to remove rule by id. It is used in case rule needs to be removed even if it forwards to the list of VSI. Allow removing all advanced rules from a particular VSI. It is useful in rebuilding VSI path. Co-developed-by: Dan Nowlin <[email protected]> Signed-off-by: Dan Nowlin <[email protected]> Signed-off-by: Shivanshu Shukla <[email protected]> Signed-off-by: Wojciech Drewek <[email protected]> Tested-by: Sandeep Penigalapati <[email protected]> Signed-off-by: Tony Nguyen <[email protected]>
2021-10-11ice: allow adding advanced rulesGrishma Kotecha4-1/+847
Define dummy packet headers to allow adding advanced rules in HW. This header is used as admin queue command parameter for adding a rule. The firmware will extract correct fields and will use them in look ups. Define each supported packets header and offsets to words used in recipe. Supported headers: - MAC + IPv4 + UDP - MAC + VLAN + IPv4 + UDP - MAC + IPv4 + TCP - MAC + VLAN + IPv4 + TCP - MAC + IPv6 + UDP - MAC + VLAN + IPv6 + UDP - MAC + IPv6 + TCP - MAC + VLAN + IPv6 + TCP Add code for creating an advanced rule. Rule needs to match defined dummy packet, if not return error, which means that this type of rule isn't currently supported. The first step in adding advanced rule is searching for an advanced recipe matching this kind of rule. If it doesn't exist new recipe is created. Dummy packet has to be filled with the correct header field value from the rule definition. It will be used to do look up in HW. Support searching for existing advance rule entry. It is used in case of adding the same rule on different VSI. In this case, instead of creating new rule, the existing one should be updated with refreshed VSI list. Add initialization for prof_res_bm_init flag to zero so that the possible resource for fv in the files can be initialized. Co-developed-by: Dan Nowlin <[email protected]> Signed-off-by: Dan Nowlin <[email protected]> Signed-off-by: Grishma Kotecha <[email protected]> Signed-off-by: Wojciech Drewek <[email protected]> Tested-by: Sandeep Penigalapati <[email protected]> Signed-off-by: Tony Nguyen <[email protected]>
2021-10-11ice: create advanced switch recipeDan Nowlin5-22/+1412
These changes introduce code for creating advanced recipes for the switch in hardware. There are a couple of recipes already defined in the HW. They apply to matching on basic protocol headers, like MAC, VLAN, MACVLAN, ethertype or direction (promiscuous), etc.. If the user wants to match on other protocol headers (eg. ip address, src/dst port etc.) or different variation of already supported protocols, there is a need to create new, more complex recipe. That new recipe is referred as 'advanced recipe', and the filtering rule created on top of that recipe is called 'advanced rule'. One recipe can have up to 5 words, but the first word is always reserved for match on switch id, so the driver can define up to 4 words for one recipe. To support recipes with more words up to 5 recipes can be chained, so 20 words can be programmed for look up. Input for adding recipe function is a list of protocols to support. Based on this list correct profile is being chosen. Correct profile means that it contains all protocol types from a list. Each profile have up to 48 field vector words and each of this word have protocol id and offset. These two fields need to match with input data for adding recipe function. If the correct profile can't be found the function returns an error. The next step after finding the correct profile is grouping words into groups. One group can have up to 4 words. This is done to simplify sending recipes to HW (because recipe also can have up to 4 words). In case of chaining (so when look up consists of more than 4 words) last recipe will always have results from the previous recipes used as words. A recipe to profile map is used to store information about which profile is associate with this recipe. This map is an array of 64 elements (max number of recipes) and each element is a 256 bits bitmap (max number of profiles) Profile to recipe map is used to store information about which recipe is associate with this profile. This map is an array of 256 elements (max number of profiles) and each element is a 64 bits bitmap (max number of recipes) Signed-off-by: Dan Nowlin <[email protected]> Signed-off-by: Wojciech Drewek <[email protected]> Tested-by: Sandeep Penigalapati <[email protected]> Signed-off-by: Tony Nguyen <[email protected]>
2021-10-11ice: manage profiles and field vectorsDan Nowlin5-1/+268
Implement functions to manage profiles and field vectors in hardware. In hardware, there are up to 256 profiles and each of these profiles can have 48 field vector words. Each field vector word is described by protocol id and offset in the packet. To add a new recipe all used profiles need to be searched. If the profile contains all required protocol ids and offsets from the recipe it can be used. The driver has to add this profile to recipe association to tell hardware that newly added recipe is going to be associated with this profile. The amount of used profiles depend on the package. To avoid searching across not used profile, max profile id value is calculated at init flow. The profile is considered as unused when all field vector words in the profile are invalid (protocol id 0xff and offset 0x1ff). Profiles are read from the package section ICE_SID_FLD_VEC_SW. Empty field vector words can be used for recipe results. Store all unused field vector words in prof_res_bm. It is a 256 elements array (max number of profiles) each element is a 48 bit bitmap (max number of field vector words). For now, support only non-tunnel profiles type. Co-developed-by: Grishma Kotecha <[email protected]> Signed-off-by: Grishma Kotecha <[email protected]> Signed-off-by: Dan Nowlin <[email protected]> Signed-off-by: Wojciech Drewek <[email protected]> Tested-by: Sandeep Penigalapati <[email protected]> Signed-off-by: Tony Nguyen <[email protected]>
2021-10-11ice: implement low level recipes functionsGrishma Kotecha2-0/+212
Add code to manage recipes and profiles on admin queue layer. Allow the driver to add a new recipe and update an existing one. Get a recipe and get a recipe to profile association is mostly used in update existing recipes code. Only default recipes can be updated. An update is done by reading recipes from HW, changing their params and calling add recipe command. Support following admin queue commands: - ice_aqc_opc_add_recipe (0x0290) - create a recipe with protocol header information and other details that determine how this recipe filter works - ice_aqc_opc_recipe_to_profile (0x0291) - associate a switch recipe to a profile - ice_aqc_opc_get_recipe (0x0292) - get details of an existing recipe - ice_aqc_opc_get_recipe_to_profile (0x0293) - get a recipe associated with profile ID Define ICE_AQC_RES_TYPE_RECIPE resource type to hold a switch recipe. It is needed when a new switch recipe needs to be created. Co-developed-by: Dan Nowlin <[email protected]> Signed-off-by: Dan Nowlin <[email protected]> Signed-off-by: Grishma Kotecha <[email protected]> Signed-off-by: Wojciech Drewek <[email protected]> Tested-by: Sandeep Penigalapati <[email protected]> Signed-off-by: Tony Nguyen <[email protected]>
2021-10-08Merge branch '100GbE' of ↵David S. Miller25-94/+1899
git://git.kernel.org/pub/scm/linux/kernel/git/tnguy/next-queue Tony Nguyen says: ==================== 100GbE Intel Wired LAN Driver Updates 2021-10-07 Michal Swiatkowski says: The following patch series introduces basic switchdev model support in ice driver. Implement the following blocks of switchdev framework: - VF port representors creation - control plane VSI definition - exception path (a. k. a. "slow-path") - to allow a virtual switch or linux bridge to receive any packet that doesn't match any hw filter - link state management of virtual ports - query virtual port statistics Hardware offload support in switchdev mode is out of scope of this patchset. Devlink interface is used to toggle between switchdev and legacy (the default) modes of the driver. --- Note: This series includes the use enum ice_status, however, we have patches in our queue to remove it from the driver [1]. We are working through the patches that precede the removal series. [1] https://patchwork.ozlabs.org/project/intel-wired-lan/list/?series=265957 ==================== Signed-off-by: David S. Miller <[email protected]>
2021-10-07Merge git://git.kernel.org/pub/scm/linux/kernel/git/netdev/netJakub Kicinski2-3/+3
No conflicts. Signed-off-by: Jakub Kicinski <[email protected]>
2021-10-07ice: add port representor ethtool ops and statsWojciech Drewek6-7/+118
Introduce the following ethtool operations for VF's representor: -get_drvinfo -get_strings -get_ethtool_stats -get_sset_count -get_link In all cases, existing operations were used with minor changes which allow us to detect if ethtool op was called for representor. Only VF VSI stats will be available for representor. Implement ndo_get_stats64 for port representor. This will update VF VSI stats and read them. Signed-off-by: Wojciech Drewek <[email protected]> Tested-by: Sandeep Penigalapati <[email protected]> Signed-off-by: Tony Nguyen <[email protected]>
2021-10-07ice: switchdev slow pathGrzegorz Nitka6-1/+156
Slow path means allowing packet to go from uplink to representor and from representor to correct VF on Rx site and from VF to representor and to uplink on Tx site. To accomplish this driver, has to set correct Tx descriptor. When packet is sent from representor to VF, destination should be set to VF VSI. When packet is sent from uplink port destination should be uplink to bypass switch infrastructure and send packet outside. On Rx site driver should check source VSI field from Rx descriptor and based on that forward packed to correct netdev. To allow this there is a target netdevs table in control plane VSI struct. Co-developed-by: Michal Swiatkowski <[email protected]> Signed-off-by: Michal Swiatkowski <[email protected]> Signed-off-by: Grzegorz Nitka <[email protected]> Tested-by: Sandeep Penigalapati <[email protected]> Signed-off-by: Tony Nguyen <[email protected]>
2021-10-07ice: rebuild switchdev when resetting all VFsGrzegorz Nitka6-1/+130
As resetting all VFs behaves mostly like creating new VFs also eswitch infrastructure has to be recreated. The easiest way to do that is to rebuild eswitch after resetting VFs. Implement helper functions to start and stop all representors queues. This is used to disable traffic on port representors. In rebuild path: - NAPI has to be disabled - eswitch environment has to be set up - new port representors have to be created, because the old one had pointer to not existing VFs - new control plane VSI ring should be remapped - NAPI hast to be enabled - rxdid has to be set to FLEX_NIC_2, because this descriptor id support source_vsi, which is needed on control plane VSI queues - port representors queues have to be started Signed-off-by: Grzegorz Nitka <[email protected]> Tested-by: Sandeep Penigalapati <[email protected]> Signed-off-by: Tony Nguyen <[email protected]>
2021-10-07ice: enable/disable switchdev when managing VFsGrzegorz Nitka4-2/+63
Only way to enable switchdev is to create VFs when the eswitch mode is set to switchdev. Check if correct mode is set and enable switchdev in function which creating VFs. Disable switchdev when user change number of VFs to 0. Changing eswitch mode back to legacy when VFs are created in switchdev mode isn't allowed. As switchdev takes care of managing filter rules, adding new rules on VF is blocked. In case of resetting VF driver has to update pointer in ice_repr struct, because after reset VSI related things can change. Co-developed-by: Wojciech Drewek <[email protected]> Signed-off-by: Wojciech Drewek <[email protected]> Signed-off-by: Grzegorz Nitka <[email protected]> Tested-by: Sandeep Penigalapati <[email protected]> Signed-off-by: Tony Nguyen <[email protected]>
2021-10-07ice: introduce new type of VSI for switchdevGrzegorz Nitka6-3/+92
New type of VSI has to be defined for switchdev control plane VSI. Number of allocated Tx and Rx queue has to be equal to amount of VFs, because each port representor should have one Tx and Rx queue. Also to not increase number of used irqs too much, control plane VSI uses only one q_vector and handle all queues in one irq. To allow handling all queues in one irq , new function to clean msix for eswitch was introduced. This function will schedule napi for each representor instead of scheduling it only for one like in normal clean irq function. Only one additional msix has to be requested. Always try to request it in ice_ena_msix_range function. Signed-off-by: Grzegorz Nitka <[email protected]> Tested-by: Sandeep Penigalapati <[email protected]> Signed-off-by: Tony Nguyen <[email protected]>
2021-10-07ice: set and release switchdev environmentGrzegorz Nitka6-13/+453
Switchdev environment has to be set up when user create VFs and eswitch mode is switchdev. Release is done when user delete all VFs. Data path in this implementation is based on control plane VSI. This VSI is used to pass traffic from port representors to corresponding VFs and vice versa. Default TX rule has to be added to forward packet to control plane VSI. This will redirect packets from VFs which don't match other rules to control plane VSI. On RX side default rule is added on uplink VSI to receive all traffic that doesn't match other rules. When setting switchdev environment all other rules from VFs should be removed. Packet to VFs will be forwarded by control plane VSI. As VF without any mac rules can't send any packet because of antispoof mechanism, VSI antispoof should be turned off on each VFs. To send packet from representor to correct VSI, destination VSI field in TX descriptor will have to be filled. Allow that by setting destination override bit in control plane VSI security config. Packet from VFs will be received on control plane VSI. Driver should decide to which netdev forward the packet. Decision is made based on src_vsi field from descriptor. There is a target netdev list in control plane VSI struct which choose netdev based on src_vsi number. Co-developed-by: Michal Swiatkowski <[email protected]> Signed-off-by: Michal Swiatkowski <[email protected]> Signed-off-by: Grzegorz Nitka <[email protected]> Tested-by: Sandeep Penigalapati <[email protected]> Signed-off-by: Tony Nguyen <[email protected]>
2021-10-07ice: allow changing lan_en and lb_en on dflt rulesMichal Swiatkowski4-1/+94
There is no way to change default lan_en and lb_en flags while adding new rule. Add function that allows changing these flags on ICE_SW_LKUP_DFLT recipe and any rule id. lan_en allows packet to go outside if rule is matched. Clearing this bit will block packet from sending it outside. lb_en allows packet to be forwarded to other VSI. Clearing this bit will block packet from forwarding it to other VSI. Signed-off-by: Michal Swiatkowski <[email protected]> Tested-by: Sandeep Penigalapati <[email protected]> Signed-off-by: Tony Nguyen <[email protected]>
2021-10-07ice: manage VSI antispoof and destination overrideMichal Swiatkowski2-0/+73
Implement functions to make setting VSI security config easier. Main function ice_update_security fills security section field and checks against error in updating VSI. Reset functions are responsible for correct filling config according to user expectations. This helper is needed because destination override is located in this section. Driver has to set this bit to allow strering Tx packet on VSI based on value in Tx descriptors. Signed-off-by: Michal Swiatkowski <[email protected]> Tested-by: Sandeep Penigalapati <[email protected]> Signed-off-by: Tony Nguyen <[email protected]>
2021-10-07ice: allow process VF opcodes in different waysMichal Swiatkowski3-37/+225
In switchdev driver shouldn't add MAC, VLAN and promisc filters on iavf demand but should return success to not break normal iavf flow. Achieve that by creating table of functions pointer with default functions used to parse iavf command. While parse iavf command, call correct function from table instead of calling function direct. When port representors are being created change functions in table to new one that behaves correctly for switchdev puprose (ignoring new filters). Change back to default ops when representors are being removed. Co-developed-by: Wojciech Drewek <[email protected]> Signed-off-by: Wojciech Drewek <[email protected]> Signed-off-by: Michal Swiatkowski <[email protected]> Tested-by: Sandeep Penigalapati <[email protected]> Signed-off-by: Tony Nguyen <[email protected]>
2021-10-07ice: introduce VF port representorMichal Swiatkowski6-2/+286
Port representor is used to manage VF from host side. To allow it each created representor registers netdevice with random hw address. Also devlink port is created for all representors. Port representor name is created based on switch id or managed by devlink core if devlink port was registered with success. Open and stop ndo ops are implemented to allow managing the VF link state. Link state is tracked in VF struct. Struct ice_netdev_priv is extended by pointer to representor field. This is needed to get correct representor from netdev struct mostly used in ndo calls. Implement helper functions to check if given netdev is netdev of port representor (ice_is_port_repr_netdev) and to get representor from netdev (ice_netdev_to_repr). As driver mostly will create or destroy port representors on all VFs instead of on single one, write functions to add and remove representor for each VF. Representor struct contains pointer to source VSI, which is VSI configured on VF, backpointer to VF, backpointer to netdev, q_vector pointer and metadata_dst which will be used in data path. Co-developed-by: Grzegorz Nitka <[email protected]> Signed-off-by: Grzegorz Nitka <[email protected]> Signed-off-by: Michal Swiatkowski <[email protected]> Tested-by: Sandeep Penigalapati <[email protected]> Signed-off-by: Tony Nguyen <[email protected]>
2021-10-07ice: Move devlink port to PF/VF structWojciech Drewek7-37/+103
Keeping devlink port inside VSI data structure causes some issues. Since VF VSI is released during reset that means that we have to unregister devlink port and register it again every time reset is triggered. With the new changes in devlink API it might cause deadlock issues. After calling devlink_port_register/devlink_port_unregister devlink API is going to lock rtnl_mutex. It's an issue when VF reset is triggered in netlink operation context (like setting VF MAC address or VLAN), because rtnl_lock is already taken by netlink. Another call of rtnl_lock from devlink API results in dead-lock. By moving devlink port to PF/VF we avoid creating/destroying it during reset. Since this patch, devlink ports are created during ice_probe, destroyed during ice_remove for PF and created during ice_repr_add, destroyed during ice_repr_rem for VF. Signed-off-by: Wojciech Drewek <[email protected]> Tested-by: Sandeep Penigalapati <[email protected]> Signed-off-by: Tony Nguyen <[email protected]>
2021-10-07ice: support basic E-Switch mode controlMichal Swiatkowski6-0/+116
Write set and get eswitch mode functions used by devlink ops. Use new pf struct member eswitch_mode to track current eswitch mode in driver. Changing eswitch mode is only allowed when there are no VFs created. Create new file for eswitch related code. Add config flag ICE_SWITCHDEV to allow user to choose if switchdev support should be enabled or disabled. Use case examples: - show current eswitch mode ('legacy' is the default one) [root@localhost]# devlink dev eswitch show pci/0000:03:00.1 pci/0000:03:00.1: mode legacy - move to 'switchdev' mode [root@localhost]# devlink dev eswitch set pci/0000:03:00.1 mode switchdev [root@localhost]# devlink dev eswitch show pci/0000:03:00.1 pci/0000:03:00.1: mode switchdev - create 2 VFs [root@localhost]# echo 2 > /sys/class/net/ens4f1/device/sriov_numvfs - unsuccessful attempt to change eswitch mode while VFs are created [root@localhost]# devlink dev eswitch set pci/0000:03:00.1 mode legacy devlink answers: Operation not supported - destroy VFs [root@localhost]# echo 0 > /sys/class/net/ens4f1/device/sriov_numvfs - restore 'legacy' mode [root@localhost]# devlink dev eswitch set pci/0000:03:00.1 mode legacy [root@localhost]# devlink dev eswitch show pci/0000:03:00.1 pci/0000:03:00.1: mode legacy Co-developed-by: Grzegorz Nitka <[email protected]> Signed-off-by: Grzegorz Nitka <[email protected]> Signed-off-by: Michal Swiatkowski <[email protected]> Tested-by: Sandeep Penigalapati <[email protected]> Signed-off-by: Tony Nguyen <[email protected]>
2021-10-06iavf: fix double unlock of crit_lockStefan Assmann1-1/+0
The crit_lock mutex could be unlocked twice as reported here https://lists.osuosl.org/pipermail/intel-wired-lan/Week-of-Mon-20210823/025525.html Remove the superfluous unlock. Technically the problem was already present before 5ac49f3c2702 as that commit only replaced the locking primitive, but no functional change. Reported-by: Dan Carpenter <[email protected]> Fixes: 5ac49f3c2702 ("iavf: use mutexes for locking of critical sections") Fixes: bac8486116b0 ("iavf: Refactor the watchdog state machine") Signed-off-by: Stefan Assmann <[email protected]> Signed-off-by: Tony Nguyen <[email protected]>
2021-10-06i40e: Fix freeing of uninitialized misc IRQ vectorSylwester Dziedziuch1-1/+2
When VSI set up failed in i40e_probe() as part of PF switch set up driver was trying to free misc IRQ vectors in i40e_clear_interrupt_scheme and produced a kernel Oops: Trying to free already-free IRQ 266 WARNING: CPU: 0 PID: 5 at kernel/irq/manage.c:1731 __free_irq+0x9a/0x300 Workqueue: events work_for_cpu_fn RIP: 0010:__free_irq+0x9a/0x300 Call Trace: ? synchronize_irq+0x3a/0xa0 free_irq+0x2e/0x60 i40e_clear_interrupt_scheme+0x53/0x190 [i40e] i40e_probe.part.108+0x134b/0x1a40 [i40e] ? kmem_cache_alloc+0x158/0x1c0 ? acpi_ut_update_ref_count.part.1+0x8e/0x345 ? acpi_ut_update_object_reference+0x15e/0x1e2 ? strstr+0x21/0x70 ? irq_get_irq_data+0xa/0x20 ? mp_check_pin_attr+0x13/0xc0 ? irq_get_irq_data+0xa/0x20 ? mp_map_pin_to_irq+0xd3/0x2f0 ? acpi_register_gsi_ioapic+0x93/0x170 ? pci_conf1_read+0xa4/0x100 ? pci_bus_read_config_word+0x49/0x70 ? do_pci_enable_device+0xcc/0x100 local_pci_probe+0x41/0x90 work_for_cpu_fn+0x16/0x20 process_one_work+0x1a7/0x360 worker_thread+0x1cf/0x390 ? create_worker+0x1a0/0x1a0 kthread+0x112/0x130 ? kthread_flush_work_fn+0x10/0x10 ret_from_fork+0x1f/0x40 The problem is that at that point misc IRQ vectors were not allocated yet and we get a call trace that driver is trying to free already free IRQ vectors. Add a check in i40e_clear_interrupt_scheme for __I40E_MISC_IRQ_REQUESTED PF state before calling i40e_free_misc_vector. This state is set only if misc IRQ vectors were properly initialized. Fixes: c17401a1dd21 ("i40e: use separate state bit for miscellaneous IRQ setup") Reported-by: PJ Waskiewicz <[email protected]> Signed-off-by: Sylwester Dziedziuch <[email protected]> Signed-off-by: Mateusz Palczewski <[email protected]> Tested-by: Dave Switzer <[email protected]> Signed-off-by: Tony Nguyen <[email protected]>
2021-10-06i40e: fix endless loop under rtnlJiri Benc1-1/+1
The loop in i40e_get_capabilities can never end. The problem is that although i40e_aq_discover_capabilities returns with an error if there's a firmware problem, the returned error is not checked. There is a check for pf->hw.aq.asq_last_status but that value is set to I40E_AQ_RC_OK on most firmware problems. When i40e_aq_discover_capabilities encounters a firmware problem, it will encounter the same problem on its next invocation. As the result, the loop becomes endless. We hit this with I40E_ERR_ADMIN_QUEUE_TIMEOUT but looking at the code, it can happen with a range of other firmware errors. I don't know what the correct behavior should be: whether the firmware should be retried a few times, or whether pf->hw.aq.asq_last_status should be always set to the encountered firmware error (but then it would be pointless and can be just replaced by the i40e_aq_discover_capabilities return value). However, the current behavior with an endless loop under the rtnl mutex(!) is unacceptable and Intel has not submitted a fix, although we explained the bug to them 7 months ago. This may not be the best possible fix but it's better than hanging the whole system on a firmware bug. Fixes: 56a62fc86895 ("i40e: init code and hardware support") Tested-by: Stefan Assmann <[email protected]> Signed-off-by: Jiri Benc <[email protected]> Reviewed-by: Jesse Brandeburg <[email protected]> Tested-by: Dave Switzer <[email protected]> Signed-off-by: Tony Nguyen <[email protected]>
2021-10-05ethernet: use eth_hw_addr_set() for dev->addr_len casesJakub Kicinski9-18/+16
Convert all Ethernet drivers from memcpy(... dev->addr_len) to eth_hw_addr_set(): @@ expression dev, np; @@ - memcpy(dev->dev_addr, np, dev->addr_len) + eth_hw_addr_set(dev, np) In theory addr_len may not be ETH_ALEN, but we don't expect non-Ethernet devices to live under this directory, and only the following cases of setting addr_len exist: - cxgb4 for mgmt device, and the drivers which set it to ETH_ALEN: s2io, mlx4, vxge. Signed-off-by: Jakub Kicinski <[email protected]> Signed-off-by: David S. Miller <[email protected]>
2021-10-02ethernet: use eth_hw_addr_set() - castsJakub Kicinski1-1/+1
eth_hw_addr_set() takes a u8 pointer, like other etherdevice helpers. Convert the few drivers which require casts because they memcpy from "endian marked" types. Signed-off-by: Jakub Kicinski <[email protected]> Signed-off-by: David S. Miller <[email protected]>
2021-10-02ethernet: use eth_hw_addr_set() instead of ether_addr_copy()Jakub Kicinski7-13/+13
Convert Ethernet from ether_addr_copy() to eth_hw_addr_set(): @@ expression dev, np; @@ - ether_addr_copy(dev->dev_addr, np) + eth_hw_addr_set(dev, np) Signed-off-by: Jakub Kicinski <[email protected]> Signed-off-by: David S. Miller <[email protected]>
2021-10-01Merge https://git.kernel.org/pub/scm/linux/kernel/git/bpf/bpf-nextJakub Kicinski3-87/+73
Daniel Borkmann says: ==================== bpf-next 2021-10-02 We've added 85 non-merge commits during the last 15 day(s) which contain a total of 132 files changed, 13779 insertions(+), 6724 deletions(-). The main changes are: 1) Massive update on test_bpf.ko coverage for JITs as preparatory work for an upcoming MIPS eBPF JIT, from Johan Almbladh. 2) Add a batched interface for RX buffer allocation in AF_XDP buffer pool, with driver support for i40e and ice from Magnus Karlsson. 3) Add legacy uprobe support to libbpf to complement recently merged legacy kprobe support, from Andrii Nakryiko. 4) Add bpf_trace_vprintk() as variadic printk helper, from Dave Marchevsky. 5) Support saving the register state in verifier when spilling <8byte bounded scalar to the stack, from Martin Lau. 6) Add libbpf opt-in for stricter BPF program section name handling as part of libbpf 1.0 effort, from Andrii Nakryiko. 7) Add a document to help clarifying BPF licensing, from Alexei Starovoitov. 8) Fix skel_internal.h to propagate errno if the loader indicates an internal error, from Kumar Kartikeya Dwivedi. 9) Fix build warnings with -Wcast-function-type so that the option can later be enabled by default for the kernel, from Kees Cook. 10) Fix libbpf to ignore STT_SECTION symbols in legacy map definitions as it otherwise errors out when encountering them, from Toke Høiland-Jørgensen. 11) Teach libbpf to recognize specialized maps (such as for perf RB) and internally remove BTF type IDs when creating them, from Hengqi Chen. 12) Various fixes and improvements to BPF selftests. ==================== Link: https://lore.kernel.org/r/[email protected] Signed-off-by: Jakub Kicinski <[email protected]>
2021-09-30Merge git://git.kernel.org/pub/scm/linux/kernel/git/netdev/netJakub Kicinski3-10/+22
drivers/net/phy/bcm7xxx.c d88fd1b546ff ("net: phy: bcm7xxx: Fixed indirect MMD operations") f68d08c437f9 ("net: phy: bcm7xxx: Add EPHY entry for 72165") net/sched/sch_api.c b193e15ac69d ("net: prevent user from passing illegal stab size") 69508d43334e ("net_sched: Use struct_size() and flex_array_size() helpers") Both cases trivial - adjacent code additions. Signed-off-by: Jakub Kicinski <[email protected]>
2021-09-30ixgbe: let the xdpdrv work with more than 64 cpusJason Xing5-24/+77
Originally, ixgbe driver doesn't allow the mounting of xdpdrv if the server is equipped with more than 64 cpus online. So it turns out that the loading of xdpdrv causes the "NOMEM" failure. Actually, we can adjust the algorithm and then make it work through mapping the current cpu to some xdp ring with the protect of @tx_lock. Here are some numbers before/after applying this patch with xdp-example loaded on the eth0X: As client (tx path): Before After TCP_STREAM send-64 734.14 714.20 TCP_STREAM send-128 1401.91 1395.05 TCP_STREAM send-512 5311.67 5292.84 TCP_STREAM send-1k 9277.40 9356.22 (not stable) TCP_RR send-1 22559.75 21844.22 TCP_RR send-128 23169.54 22725.13 TCP_RR send-512 21670.91 21412.56 As server (rx path): Before After TCP_STREAM send-64 1416.49 1383.12 TCP_STREAM send-128 3141.49 3055.50 TCP_STREAM send-512 9488.73 9487.44 TCP_STREAM send-1k 9491.17 9356.22 (not stable) TCP_RR send-1 23617.74 23601.60 ... Notice: the TCP_RR mode is unstable as the official document explains. I tested many times with different parameters combined through netperf. Though the result is not that accurate, I cannot see much influence on this patch. The static key is places on the hot path, but it actually shouldn't cause a huge regression theoretically. Co-developed-by: Shujin Li <[email protected]> Signed-off-by: Shujin Li <[email protected]> Signed-off-by: Jason Xing <[email protected]> Tested-by: Sandeep Penigalapati <[email protected]> Signed-off-by: Tony Nguyen <[email protected]> Signed-off-by: David S. Miller <[email protected]>
2021-09-29ixgbe: Fix NULL pointer dereference in ixgbe_xdp_setupFeng Zhou2-3/+7
The ixgbe driver currently generates a NULL pointer dereference with some machine (online cpus < 63). This is due to the fact that the maximum value of num_xdp_queues is nr_cpu_ids. Code is in "ixgbe_set_rss_queues"". Here's how the problem repeats itself: Some machine (online cpus < 63), And user set num_queues to 63 through ethtool. Code is in the "ixgbe_set_channels", adapter->ring_feature[RING_F_FDIR].limit = count; It becomes 63. When user use xdp, "ixgbe_set_rss_queues" will set queues num. adapter->num_rx_queues = rss_i; adapter->num_tx_queues = rss_i; adapter->num_xdp_queues = ixgbe_xdp_queues(adapter); And rss_i's value is from f = &adapter->ring_feature[RING_F_FDIR]; rss_i = f->indices = f->limit; So "num_rx_queues" > "num_xdp_queues", when run to "ixgbe_xdp_setup", for (i = 0; i < adapter->num_rx_queues; i++) if (adapter->xdp_ring[i]->xsk_umem) It leads to panic. Call trace: [exception RIP: ixgbe_xdp+368] RIP: ffffffffc02a76a0 RSP: ffff9fe16202f8d0 RFLAGS: 00010297 RAX: 0000000000000000 RBX: 0000000000000020 RCX: 0000000000000000 RDX: 0000000000000000 RSI: 000000000000001c RDI: ffffffffa94ead90 RBP: ffff92f8f24c0c18 R8: 0000000000000000 R9: 0000000000000000 R10: ffff9fe16202f830 R11: 0000000000000000 R12: ffff92f8f24c0000 R13: ffff9fe16202fc01 R14: 000000000000000a R15: ffffffffc02a7530 ORIG_RAX: ffffffffffffffff CS: 0010 SS: 0018 7 [ffff9fe16202f8f0] dev_xdp_install at ffffffffa89fbbcc 8 [ffff9fe16202f920] dev_change_xdp_fd at ffffffffa8a08808 9 [ffff9fe16202f960] do_setlink at ffffffffa8a20235 10 [ffff9fe16202fa88] rtnl_setlink at ffffffffa8a20384 11 [ffff9fe16202fc78] rtnetlink_rcv_msg at ffffffffa8a1a8dd 12 [ffff9fe16202fcf0] netlink_rcv_skb at ffffffffa8a717eb 13 [ffff9fe16202fd40] netlink_unicast at ffffffffa8a70f88 14 [ffff9fe16202fd80] netlink_sendmsg at ffffffffa8a71319 15 [ffff9fe16202fdf0] sock_sendmsg at ffffffffa89df290 16 [ffff9fe16202fe08] __sys_sendto at ffffffffa89e19c8 17 [ffff9fe16202ff30] __x64_sys_sendto at ffffffffa89e1a64 18 [ffff9fe16202ff38] do_syscall_64 at ffffffffa84042b9 19 [ffff9fe16202ff50] entry_SYSCALL_64_after_hwframe at ffffffffa8c0008c So I fix ixgbe_max_channels so that it will not allow a setting of queues to be higher than the num_online_cpus(). And when run to ixgbe_xdp_setup, take the smaller value of num_rx_queues and num_xdp_queues. Fixes: 4a9b32f30f80 ("ixgbe: fix potential RX buffer starvation for AF_XDP") Signed-off-by: Feng Zhou <[email protected]> Tested-by: Sandeep Penigalapati <[email protected]> Signed-off-by: Tony Nguyen <[email protected]> Signed-off-by: David S. Miller <[email protected]>
2021-09-28ice: Prefer kcalloc over open coded arithmeticLen Baker1-1/+1
As noted in the "Deprecated Interfaces, Language Features, Attributes, and Conventions" documentation [1], size calculations (especially multiplication) should not be performed in memory allocator (or similar) function arguments due to the risk of them overflowing. This could lead to values wrapping around and a smaller allocation being made than the caller was expecting. Using those allocations could lead to linear overflows of heap memory and other misbehaviors. In this case this is not actually dynamic sizes: both sides of the multiplication are constant values. However it is best to refactor this anyway, just to keep the open-coded math idiom out of code. So, use the purpose specific kcalloc() function instead of the argument size * count in the kzalloc() function. [1] https://www.kernel.org/doc/html/v5.14/process/deprecated.html#open-coded-arithmetic-in-allocator-arguments Signed-off-by: Len Baker <[email protected]> Reviewed-by: Gustavo A. R. Silva <[email protected]> Signed-off-by: Tony Nguyen <[email protected]>
2021-09-28ice: Fix macro name for IPv4 fragment flagJeff Guo2-2/+2
In IPv4 header, fragment flags indicate whether the packet needs to be fragmented or not. The value 0x20 represents MF (More Fragment); fix the macro name to match this. Signed-off-by: Ting Xu <[email protected]> Signed-off-by: Jeff Guo <[email protected]> Tested-by: Gurucharan G <[email protected]> Signed-off-by: Tony Nguyen <[email protected]>
2021-09-28ice: refactor devlink getter/fallback functions to voidJacob Keller1-83/+50
After commit a8f89fa27773 ("ice: do not abort devlink info if board identifier can't be found"), the getter/fallback() functions no longer report an error. Convert the interface to a void so that it is no longer possible to add a version field that is fatal. This makes sense, because we should not fail to report other versions just because one of the version pieces could not be found. Finally, clean up the getter functions line wrapping so that none of them take more than 80 columns, as is the usual style for networking files. Signed-off-by: Jacob Keller <[email protected]> Tested-by: Tony Brelinski <[email protected]> Signed-off-by: Tony Nguyen <[email protected]>
2021-09-28ice: Fix link mode handlingAnirudh Venkataramanan1-2/+6
The messaging for unsupported module detection is different for lenient mode and strict mode. Update the code to print the right messaging for a given link mode. Media topology conflict is not an error in lenient mode, so return an error code only if not in lenient mode. Signed-off-by: Anirudh Venkataramanan <[email protected]> Tested-by: Gurucharan G <[email protected]> Signed-off-by: Tony Nguyen <[email protected]>
2021-09-28ice: Add feature bitmap, helpers and a check for DSCPAnirudh Venkataramanan5-2/+63
DSCP a.k.a L3 QoS is only supported on certain devices. To enforce this, this patch introduces a bitmap of features and helper functions. The feature bitmap is set based on device IDs on driver init. Currently, DSCP is the only feature in this bitmap, but there will be more in the future. In the DCB netlink flow, check if the feature bit is set before exercising DSCP. Signed-off-by: Anirudh Venkataramanan <[email protected]> Tested-by: Gurucharan G <[email protected]> Signed-off-by: Tony Nguyen <[email protected]>
2021-09-28ice: Add DSCP supportDave Ertman11-56/+464
Implement code to handle submission of APP TLV's containing DSCP to TC mapping. The first such mapping received on an interface will cause that PF to switch to L3 DSCP QoS mode, apply the default config for that mode, and apply the received mapping. Only one such mapping will be allowed per DSCP value, and when the last DSCP mapping is deleted, the PF will switch back into L2 VLAN QoS mode, applying the appropriate default QoS settings. L3 DSCP QoS mode will only be allowed in SW DCBx mode, in other words, when the FW LLDP engine is disabled. Commands that break this mutual exclusivity will be blocked. Co-developed-by: Anirudh Venkataramanan <[email protected]> Signed-off-by: Anirudh Venkataramanan <[email protected]> Signed-off-by: Dave Ertman <[email protected]> Tested-by: Gurucharan G <[email protected]> Signed-off-by: Tony Nguyen <[email protected]>
2021-09-28i40e: Use the xsk batched rx allocation interfaceMagnus Karlsson1-27/+25
Use the new xsk batched rx allocation interface for the zero-copy data path. As the array of struct xdp_buff pointers kept by the driver is really a ring that wraps, the allocation routine is modified to detect a wrap and in that case call the allocation function twice. The allocation function cannot deal with wrapped rings, only arrays. As we now know exactly how many buffers we get and that there is no wrapping, the allocation function can be simplified even more as all if-statements in the allocation loop can be removed, improving performance. Signed-off-by: Magnus Karlsson <[email protected]> Signed-off-by: Daniel Borkmann <[email protected]> Link: https://lore.kernel.org/bpf/[email protected]
2021-09-28ice: Use the xsk batched rx allocation interfaceMagnus Karlsson1-25/+19
Use the new xsk batched rx allocation interface for the zero-copy data path. As the array of struct xdp_buff pointers kept by the driver is really a ring that wraps, the allocation routine is modified to detect a wrap and in that case call the allocation function twice. The allocation function cannot deal with wrapped rings, only arrays. As we now know exactly how many buffers we get and that there is no wrapping, the allocation function can be simplified even more as all if-statements in the allocation loop can be removed, improving performance. Signed-off-by: Magnus Karlsson <[email protected]> Signed-off-by: Daniel Borkmann <[email protected]> Link: https://lore.kernel.org/bpf/[email protected]
2021-09-28ice: Use xdp_buf instead of rx_buf for xsk zero-copyMagnus Karlsson2-39/+33
In order to use the new xsk batched buffer allocation interface, a pointer to an array of struct xsk_buff pointers need to be provided so that the function can put the result of the allocation there. In the ice driver, we already have a ring that stores pointers to xdp_buffs. This is only used for the xsk zero-copy driver and is a union with the structure that is used for the regular non zero-copy path. Unfortunately, that structure is larger than the xdp_buffs pointers which mean that there will be a stride (of 20 bytes) between each xdp_buff pointer. And feeding this into the xsk_buff_alloc_batch interface will not work since it assumes a regular array of xdp_buff pointers (each 8 bytes with 0 bytes in-between them on a 64-bit system). To fix this, remove the xdp_buff pointer from the rx_buf union and move it one step higher to the union above which only has pointers to arrays in it. This solves the problem and we can directly feed the SW ring of xdp_buff pointers straight into the allocation function in the next patch when that interface is used. This will improve performance. Signed-off-by: Magnus Karlsson <[email protected]> Signed-off-by: Daniel Borkmann <[email protected]> Link: https://lore.kernel.org/bpf/[email protected]
2021-09-27e100: fix buffer overrun in e100_get_regsJacob Keller1-6/+10
The e100_get_regs function is used to implement a simple register dump for the e100 device. The data is broken into a couple of MAC control registers, and then a series of PHY registers, followed by a memory dump buffer. The total length of the register dump is defined as (1 + E100_PHY_REGS) * sizeof(u32) + sizeof(nic->mem->dump_buf). The logic for filling in the PHY registers uses a convoluted inverted count for loop which counts from E100_PHY_REGS (0x1C) down to 0, and assigns the slots 1 + E100_PHY_REGS - i. The first loop iteration will fill in [1] and the final loop iteration will fill in [1 + 0x1C]. This is actually one more than the supposed number of PHY registers. The memory dump buffer is then filled into the space at [2 + E100_PHY_REGS] which will cause that memcpy to assign 4 bytes past the total size. The end result is that we overrun the total buffer size allocated by the kernel, which could lead to a panic or other issues due to memory corruption. It is difficult to determine the actual total number of registers here. The only 8255x datasheet I could find indicates there are 28 total MDI registers. However, we're reading 29 here, and reading them in reverse! In addition, the ethtool e100 register dump interface appears to read the first PHY register to determine if the device is in MDI or MDIx mode. This doesn't appear to be documented anywhere within the 8255x datasheet. I can only assume it must be in register 28 (the extra register we're reading here). Lets not change any of the intended meaning of what we copy here. Just extend the space by 4 bytes to account for the extra register and continue copying the data out in the same order. Change the E100_PHY_REGS value to be the correct total (29) so that the total register dump size is calculated properly. Fix the offset for where we copy the dump buffer so that it doesn't overrun the total size. Re-write the for loop to use counting up instead of the convoluted down-counting. Correct the mdio_read offset to use the 0-based register offsets, but maintain the bizarre reverse ordering so that we have the ABI expected by applications like ethtool. This requires and additional subtraction of 1. It seems a bit odd but it makes the flow of assignment into the register buffer easier to follow. Fixes: 1da177e4c3f4 ("Linux-2.6.12-rc2") Reported-by: Felicitas Hetzelt <[email protected]> Signed-off-by: Jacob Keller <[email protected]> Tested-by: Jacob Keller <[email protected]> Signed-off-by: Tony Nguyen <[email protected]>
2021-09-27e100: fix length calculation in e100_get_regs_lenJacob Keller1-1/+5
commit abf9b902059f ("e100: cleanup unneeded math") tried to simplify e100_get_regs_len and remove a double 'divide and then multiply' calculation that the e100_reg_regs_len function did. This change broke the size calculation entirely as it failed to account for the fact that the numbered registers are actually 4 bytes wide and not 1 byte. This resulted in a significant under allocation of the register buffer used by e100_get_regs. Fix this by properly multiplying the register count by u32 first before adding the size of the dump buffer. Fixes: abf9b902059f ("e100: cleanup unneeded math") Reported-by: Felicitas Hetzelt <[email protected]> Signed-off-by: Jacob Keller <[email protected]> Signed-off-by: Tony Nguyen <[email protected]>
2021-09-27ice: Open devlink when device is readyLeon Romanovsky1-4/+2
Move devlink_registration routine to be the last command, when the device is fully initialized. Signed-off-by: Leon Romanovsky <[email protected]> Signed-off-by: David S. Miller <[email protected]>
2021-09-24ice: Delete always true check of PF pointerLeon Romanovsky1-3/+0
PF pointer is always valid when PCI core calls its .shutdown() and .remove() callbacks. There is no need to check it again. Fixes: 837f08fdecbe ("ice: Add basic driver framework for Intel(R) E800 Series") Signed-off-by: Leon Romanovsky <[email protected]> Signed-off-by: David S. Miller <[email protected]>
2021-09-23Merge git://git.kernel.org/pub/scm/linux/kernel/git/netdev/netJakub Kicinski1-0/+1
net/mptcp/protocol.c 977d293e23b4 ("mptcp: ensure tx skbs always have the MPTCP ext") efe686ffce01 ("mptcp: ensure tx skbs always have the MPTCP ext") same patch merged in both trees, keep net-next. Signed-off-by: Jakub Kicinski <[email protected]>
2021-09-22devlink: Make devlink_register to be voidLeon Romanovsky3-16/+4
devlink_register() can't fail and always returns success, but all drivers are obligated to check returned status anyway. This adds a lot of boilerplate code to handle impossible flow. Make devlink_register() void and simplify the drivers that use that API call. Signed-off-by: Leon Romanovsky <[email protected]> Acked-by: Simon Horman <[email protected]> Acked-by: Vladimir Oltean <[email protected]> # dsa Reviewed-by: Jiri Pirko <[email protected]> Signed-off-by: David S. Miller <[email protected]>
2021-09-19igc: fix build errors for PTPRandy Dunlap1-0/+1
When IGC=y and PTP_1588_CLOCK=m, the ptp_*() interface family is not available to the igc driver. Make this driver depend on PTP_1588_CLOCK_OPTIONAL so that it will build without errors. Various igc commits have used ptp_*() functions without checking that PTP_1588_CLOCK is enabled. Fix all of these here. Fixes these build errors: ld: drivers/net/ethernet/intel/igc/igc_main.o: in function `igc_msix_other': igc_main.c:(.text+0x6494): undefined reference to `ptp_clock_event' ld: igc_main.c:(.text+0x64ef): undefined reference to `ptp_clock_event' ld: igc_main.c:(.text+0x6559): undefined reference to `ptp_clock_event' ld: drivers/net/ethernet/intel/igc/igc_ethtool.o: in function `igc_ethtool_get_ts_info': igc_ethtool.c:(.text+0xc7a): undefined reference to `ptp_clock_index' ld: drivers/net/ethernet/intel/igc/igc_ptp.o: in function `igc_ptp_feature_enable_i225': igc_ptp.c:(.text+0x330): undefined reference to `ptp_find_pin' ld: igc_ptp.c:(.text+0x36f): undefined reference to `ptp_find_pin' ld: drivers/net/ethernet/intel/igc/igc_ptp.o: in function `igc_ptp_init': igc_ptp.c:(.text+0x11cd): undefined reference to `ptp_clock_register' ld: drivers/net/ethernet/intel/igc/igc_ptp.o: in function `igc_ptp_stop': igc_ptp.c:(.text+0x12dd): undefined reference to `ptp_clock_unregister' ld: drivers/platform/x86/dell/dell-wmi-privacy.o: in function `dell_privacy_wmi_probe': Fixes: 64433e5bf40ab ("igc: Enable internal i225 PPS") Fixes: 60dbede0c4f3d ("igc: Add support for ethtool GET_TS_INFO command") Fixes: 87938851b6efb ("igc: enable auxiliary PHC functions for the i225") Fixes: 5f2958052c582 ("igc: Add basic skeleton for PTP") Signed-off-by: Randy Dunlap <[email protected]> Cc: Ederson de Souza <[email protected]> Cc: Tony Nguyen <[email protected]> Cc: Vinicius Costa Gomes <[email protected]> Cc: Jeff Kirsher <[email protected]> Cc: "David S. Miller" <[email protected]> Cc: Jakub Kicinski <[email protected]> Cc: Jesse Brandeburg <[email protected]> Cc: [email protected] Acked-by: Vinicius Costa Gomes <[email protected]> Signed-off-by: David S. Miller <[email protected]>
2021-09-17net: e1000e: solve insmod 'Unknown symbol mutex_lock' errorHao Chen1-0/+1
After I turn on the CONFIG_LOCK_STAT=y, insmod e1000e.ko will report: [ 5.641579] e1000e: Unknown symbol mutex_lock (err -2) [ 90.775705] e1000e: Unknown symbol mutex_lock (err -2) [ 132.252339] e1000e: Unknown symbol mutex_lock (err -2) This problem fixed after include <linux/mutex.h>. Signed-off-by: Hao Chen <[email protected]> Signed-off-by: David S. Miller <[email protected]>
2021-09-16igc: fix tunnel offloadingPaolo Abeni1-1/+3
Checking tunnel offloading, it turns out that offloading doesn't work as expected. The following script allows to reproduce the issue. Call it as `testscript DEVICE LOCALIP REMOTEIP NETMASK' === SNIP === if [ $# -ne 4 ] then echo "Usage $0 DEVICE LOCALIP REMOTEIP NETMASK" exit 1 fi DEVICE="$1" LOCAL_ADDRESS="$2" REMOTE_ADDRESS="$3" NWMASK="$4" echo "Driver: $(ethtool -i ${DEVICE} | awk '/^driver:/{print $2}') " ethtool -k "${DEVICE}" | grep tx-udp echo echo "Set up NIC and tunnel..." ip addr add "${LOCAL_ADDRESS}/${NWMASK}" dev "${DEVICE}" ip link set "${DEVICE}" up sleep 2 ip link add vxlan1 type vxlan id 42 \ remote "${REMOTE_ADDRESS}" \ local "${LOCAL_ADDRESS}" \ dstport 0 \ dev "${DEVICE}" ip addr add fc00::1/64 dev vxlan1 ip link set vxlan1 up sleep 2 rm -f vxlan.pcap echo "Running tcpdump and iperf3..." ( nohup tcpdump -i any -w vxlan.pcap >/dev/null 2>&1 ) & sleep 2 iperf3 -c fc00::2 >/dev/null pkill tcpdump echo echo -n "Max. Paket Size: " tcpdump -r vxlan.pcap -nnle 2>/dev/null \ | grep "${LOCAL_ADDRESS}.*> ${REMOTE_ADDRESS}.*OTV" \ | awk '{print $8}' | awk -F ':' '{print $1}' \ | sort -n | tail -1 echo ip link del vxlan1 ip addr del ${LOCAL_ADDRESS}/${NWMASK} dev "${DEVICE}" === SNAP === The expected outcome is Max. Paket Size: 64904 This is what you see on igb, the code igc has been taken from. However, on igc the output is Max. Paket Size: 1516 so the GSO aggregate packets are segmented by the kernel before calling igc_xmit_frame. Inside the subsequent call to igc_tso, the check for skb_is_gso(skb) fails and the function returns prematurely. It turns out that this occurs because the feature flags aren't set entirely correctly in igc_probe. In contrast to the original code from igb_probe, igc_probe neglects to set the flags required to allow tunnel offloading. Setting the same flags as igb fixes the issue on igc. Fixes: 34428dff3679 ("igc: Add GSO partial support") Signed-off-by: Paolo Abeni <[email protected]> Tested-by: Corinna Vinschen <[email protected]> Acked-by: Sasha Neftin <[email protected]> Tested-by: Nechama Kraus <[email protected]> Signed-off-by: Tony Nguyen <[email protected]> Signed-off-by: David S. Miller <[email protected]>
2021-09-10ice: Correctly deal with PFs that do not support RDMADave Ertman2-0/+8
There are two cases where the current PF does not support RDMA functionality. The first is if the NVM loaded on the device is set to not support RDMA (common_caps.rdma is false). The second is if the kernel bonding driver has included the current PF in an active link aggregate. When the driver has determined that this PF does not support RDMA, then auxiliary devices should not be created on the auxiliary bus. Without a device on the auxiliary bus, even if the irdma driver is present, there will be no RDMA activity attempted on this PF. Currently, in the reset flow, an attempt to create auxiliary devices is performed without regard to the ability of the PF. There needs to be a check in ice_aux_plug_dev (as the central point that creates auxiliary devices) to see if the PF is in a state to support the functionality. When disabling and re-enabling RDMA due to the inclusion/removal of the PF in a link aggregate, we also need to set/clear the bit which controls auxiliary device creation so that a reset recovery in a link aggregate situation doesn't try to create auxiliary devices when it shouldn't. Fixes: f9f5301e7e2d ("ice: Register auxiliary device to provide RDMA") Reported-by: Yongxin Liu <[email protected]> Signed-off-by: Dave Ertman <[email protected]> Signed-off-by: Tony Nguyen <[email protected]> Signed-off-by: David S. Miller <[email protected]>
2021-08-31Merge git://git.kernel.org/pub/scm/linux/kernel/git/netdev/netJakub Kicinski2-16/+63
include/linux/netdevice.h net/socket.c d0efb16294d1 ("net: don't unconditionally copy_from_user a struct ifreq for socket ioctls") 876f0bf9d0d5 ("net: socket: simplify dev_ifconf handling") 29c4964822aa ("net: socket: rework compat_ifreq_ioctl()") Signed-off-by: Jakub Kicinski <[email protected]>
Powered by cgit, Gruvboxed by Blaster4385.