- 17 Sep, 2023 23 commits
-
-
Jiri Pirko authored
In mlx5, there is a devlink instance created for PCI device. Also, one separate devlink instance is created for auxiliary device that represents the netdev of uplink port. This relation is currently invisible to the devlink user. Benefit from the rel infrastructure and allow for nested devlink instance to set the relationship for the nested-in devlink instance. Note that there may be many nested instances, therefore use xarray to hold the list of rel_indexes for individual nested instances. Signed-off-by:
Jiri Pirko <jiri@nvidia.com> Signed-off-by:
David S. Miller <davem@davemloft.net>
-
Jiri Pirko authored
Benefit from the newly introduced rel infrastructure, treat the linecard nested devlink instances in the same way as port function instances. Convert the code to use the rel infrastructure. Signed-off-by:
-
Jiri Pirko authored
Benefit from the existence of internal mlx5 notifier and extend it by event MLX5_DRIVER_EVENT_SF_PEER_DEVLINK. Use this event from SF auxiliary device probe/remove functions to pass the registered SF devlink instance to the SF representor. Process the new event in SF representor code and call devl_port_fn_devlink_set() to do the assignments. Implement this in work to avoid possible deadlock when probe/remove function of SF may be called with devlink instance lock held during devlink reload. Signed-off-by:
-
Jiri Pirko authored
Introduce a new helper devl_port_fn_devlink_set() to be used by driver assigning a devlink instance to the peer devlink port function. Expose this to user over new netlink attribute nested under port function nest to expose devlink handle related to the port function. This is particularly helpful for user to understand the relationship between devlink instances created for SFs and the port functions they belong to. Note that caller of devlink_port_notify() needs to hold devlink instance lock, put the assertion to devl_port_fn_devlink_set() to make this requirement explicit. Also note the limitations that only allow to make this assignment for registered objects. Signed-off-by:
-
Jiri Pirko authored
It is a bit tricky to maintain relationship between devlink objects and nested devlink instances due to following aspects: 1) Locking. It is necessary to lock the devlink instance that contains the object first, only after that to lock the nested instance. 2) Lifetimes. Objects (e.g devlink port) may be removed before the nested devlink instance. 3) Notifications. If nested instance changes (e.g. gets registered/unregistered) the nested-in object needs to send appropriate notifications. Resolve this by introducing an xarray that holds 1:1 relationships between devlink object and related nested devlink instance. Use that xarray index to get the object/nested devlink instance on the other side. Provide necessary helpers: devlink_rel_nested_in_add/clear() to add and clear the relationship. devlink_rel_nested_in_notify() to call the nested-in object to send notifications during nested instance register/unregister/netns change. devlink_rel_devlink_handle_put() to be used by nested-in object fill function to fill the nested handle. Signed-off-by:
-
Jiri Pirko authored
As the next patch is going to call this helper with need to fill another type of nested attribute, pass it over function arg. Signed-off-by:
-
Jiri Pirko authored
As the next patch is going to call this helper out of the linecard.c, move to netlink.c. Signed-off-by:
-
Jiri Pirko authored
If netns of devlink instance and nested devlink instance differs, put netnsid attr to indicate that. Signed-off-by:
-
Jiri Pirko authored
Historically, the shared devlink_mutex prevented devlink instances from being registered/unregistered during another devlink instance reload operation. However, devlink_muxex is gone for some time now, this limitation is no longer needed. Lift it. Signed-off-by:
-
Jiri Pirko authored
The eswitch disable call does removal of all representors. Do that before clearing the SF device table and maintain the same flow as during SF devlink port removal, where the representor is removed before the actual SF is removed. Signed-off-by:
-
Jiri Pirko authored
Instead of exposing linecard struct, expose a simple helper to get the linecard index, which is all is needed outside linecard.c. Move the linecard struct to linecard.c and keep it private similar to the rest of the devlink objects. Signed-off-by:
-
Pavithra Sathyanarayanan authored
When operating in fixed phy mode and if there is repeated open/close phy test cases, everytime the fixed phy is registered as a new phy which leads to overrun after 32 iterations. It is solved by adding fixed_phy_unregister() in the phy_close path. In phy_close path, netdev->phydev cannot be used directly in fixed_phy_unregister() due to two reasons, - netdev->phydev is set to NULL in phy_disconnect() - fixed_phy_unregister() can be called only after phy_disconnect() So saving the netdev->phydev in local variable 'phydev' and passing it to phy_disconnect(). Signed-off-by:Pavithra Sathyanarayanan <Pavithra.Sathyanarayanan@microchip.com> Signed-off-by:
-
David S. Miller authored
Vadim Fedorenko says: ==================== Create common DPLL configuration API Implement common API for DPLL configuration and status reporting. The API utilises netlink interface as transport for commands and event notifications. This API aims to extend current pin configuration provided by PTP subsystem and make it flexible and easy to cover complex configurations. Netlink interface is based on ynl spec, it allows use of in-kernel tools/net/ynl/cli.py application to control the interface with properly formated command and json attribute strings. Here are few command examples of how it works with `ice` driver on supported NIC: - dump dpll devices: $ sudo ./tools/net/ynl/cli.py --spec Documentation/netlink/specs/dpll.yaml \ --dump device-get [{'clock-id': 4658613174691613800, 'id': 0, 'lock-status': 'locked-ho-acq', 'mode': 'automatic', 'mode-supported': ['automatic'], 'module-name': 'ice', 'type': 'eec'}, {'clock-id': 4658613174691613800, 'id': 1, 'lock-status': 'locked-ho-acq', 'mode': 'automatic', 'mode-supported': ['automatic'], 'module-name': 'ice', 'type': 'pps'}] - get single pin info: $ sudo ./tools/net/ynl/cli.py --spec Documentation/netlink/specs/dpll.yaml \ --do pin-get --json '{"id":2}' {'board-label': 'C827_0-RCLKA', 'clock-id': 4658613174691613800, 'capabilities': 6, 'frequency': 1953125, 'id': 2, 'module-name': 'ice', 'parent-device': [{'direction': 'input', 'parent-id': 0, 'prio': 9, 'state': 'disconnected'}, {'direction': 'input', 'parent-id': 1, 'prio': 9, 'state': 'disconnected'}], 'type': 'mux'} - set pin's state on dpll: $ sudo ./tools/net/ynl/cli.py --spec Documentation/netlink/specs/dpll.yaml \ --do pin-set --json '{"id":2, "parent-device":{"parent-id":1, "state":2}}' - set pin's prio on dpll: $ sudo ./tools/net/ynl/cli.py --spec Documentation/netlink/specs/dpll.yaml \ --do pin-set --json '{"id":2, "parent-device":{"parent-id":1, "prio":4}}' - set pin's state on parent pin: $ sudo ./tools/net/ynl/cli.py --spec Documentation/netlink/specs/dpll.yaml \ --do pin-set --json '{"id":13, "parent-pin":{"parent-id":2, "state":1}}' Changelog: v7 -> v8: - rebase on top of net-next - no functional changes in patchset v6 -> v7: - use unique id in references array to prevent possible crashes v5 -> v6: - change dpll-caps to pin capabilities and adjust enum accordingly - remove dpll.h from netdevice.h v4 -> v5: - separate namespace for pin attributes - small fixes, more details in the patches v3 -> v4: - rebase on top of net-next - fix flag usage in ice v2 -> v3: - more style and warning fixes - details in per-patch logs v1 -> v2: - remove FREERUN/DETACHED mode - reorder functions in commits not to depend on files introduced in future commits - style and warning fixes v9 RFC -> v1: - Merge header patch into the patches where the actual functions are implemented - Address comments from previous reviews - Per patch change log contains more details RFC versions: v8 -> v9: [00/10] Create common DPLL configuration API - update examples to reflect new pin-parent nest split [01/10] dpll: documentation on DPLL subsystem interface - fix docs build warnings - separate netlink command/attribute list - replace enum description with uapi header - add brief explanation what is a DPLL - fix EOPNOTSUPP typo - fix typo .state_get -> .state_on_dpll_get [02/10] dpll: spec: Add Netlink spec in YAML - regenerate policy max values - add missing enum descriptions - split pin-parent nest: - pin-parent-device - for configuration of pin-device tuple - pin-parent-pin - for configuration od pin-pin tuple - fix typos: - s/working-modes/working modes/ - s/differentiate/differentiates/ - s/valid input, auto selected by dpll/input pin auto selected by dpll/ - remove FREERUN and HOLDOVER modes [03/10] dpll: core: Add DPLL framework base functions - fix description in spdx header. - remove refcount check if refcount was already set - do not validate dpll ptr in dpll_device_put(..) - fix return -ENOMEM on failed memory alloc - do not validate pin ptr in dpll_pin_put(..) - return -EINVAL in case of module/clock_id mismatch - do not {} around one-line xa_for_each() macro - move dpll_<x>_registration structs to dpll_core.c - rephrase doc comment on device and pin id struct members - remove ref in case of memory allocation fail - check for required ops on pin/device registration - mark pin with DPLL_REGISTERED once pin is registered with dpll [04/10] dpll: netlink: Add DPLL framework base functions - fix pin-id-get/device-id-get behavior - reshuffle order of functions - avoid forward declarations - functions for adding pin/device handle next to each other - pass ops callback return values to the user - remove dpll_cmd_pin_fill_details(..) function, merge the code into __dpll_cmd_pin_dump_one(..) - rename __dpll_cmd_pin_dump_one() to dpll_cmd_pin_get_one() - use WARN_ON macro when dpll ref is missing - remove redundant pin's dpll list not empty check - remove double spaces inside if statement - add extack message when set command is not possible - do not return error when callback is not required - WARN_ON missing ops moved to dpll_core.c - use DPLL_REGISTERED if pin was registered with dpll - fix pin-id-get return and add extack errors - fix device-id-get return and add extack errors - drop pointless init of variables - add macro for iterating over marked pins/devices - move dpll_set_from_nlattr() for consistent order - use GENL_REQ_ATTR_CHECK() for checking attibute presence - fill extack if pin/device was not found - drop pointless init of variables - WARN_ON if dpll not registered on send event - rename goto labels to indicate error path - fix docs - drop pointless init of variables - verify pin in notify with a mark - prevent ops->mode_set call if missing callback - move static dpll_msg_add_pin_handle() from pin<->netdev patch - split pin-parent nest: - pin-parent-device - for configuration of pin-device tuple - pin-parent-pin - for configuration od pin-pin tuple [06/10] netdev: expose DPLL pin handle for netdevice - net_device->dpll_pin is only valid if IS_ENABLED(CONFIG_DPLL) fix the code in net/core/rtnetlink.c to respect that. - move dpll_msg_add_pin_handle to "dpll: netlink" patch + export the function with this patch [07/10] ice: add admin commands to access cgu configuration - rename MAX_NETLIST_SIZE -> ICE_MAX_NETLIST_SIZE - simplify function: s64 convert_s48_to_s64(s64 signed_48) - do not assign 0 to field that is already 0 [08/10] ice: implement dpll interface to control cgu - drop pointless 0 assignement - ice_dpll_init(..) returns void instead of int - fix context description of the functions - fix ice_dpll_init(..) traces - fix use package_label instead pf board_label for rclk pin - be consistent on cgu presence naming - remove indent in ice_dpll_deinit(..) - remove unused struct field lock_err_num - fix kworker resched behavior - remove debug log from ice_dpll_deinit_worker(..) - reorder ice internal functions - release resources directly on error path - remove redundant NULL checks when releasing resources - do not assign NULL to pointers after releasing resources - simplify variable assignement - fix 'int ret;' declarations across the ice_dpll.c - remove leftover ice_dpll_find(..) - get pf pointer from dpll_priv without type cast - improve error reporting - fix documentation - fix ice_dpll_update_state(..) flow - fix return in case out of range prio set v7 -> v8: [0/10] Create common DPLL configuration API - reorder the patches in patch series - split patch "[RFC PATCH v7 2/8] dpll: Add DPLL framework base functions" into 3 smaller patches for easier review: - [03/10] dpll: core: Add DPLL framework base functions - [04/10] dpll: netlink: Add DPLL framework base functions - [05/10] dpll: api header: Add DPLL framework base - add cli.py usage examples in commit message [01/10] dpll: documentation on DPLL subsystem interface - fix DPLL_MODE_MANUAL documentation - remove DPLL_MODE_NCO - remove DPLL_LOCK_STATUS_CALIBRATING - add grepability Use full names of commands, attributes and values of dpll subsystem in the documentation - align documentation with changes introduced in v8 - fix typos - fix phrases to better show the intentions - move dpll.rst to Documentation/driver-api/ [02/10] dpll: spec: Add Netlink spec in YAML - remove unspec attribute values - add 10 KHZ and 77,5 KHZ frequency defines - fix documentation - remove assigned values from subset attributes - reorder dpll attributes - fix `device` nested attribute usage, device get is not used on pin-get - temperature with 3 digit float precision - remove enum from subset definitions - move pin-direction to pin-dpll tuple/subset - remove DPLL_MODE_NCO - remove DPLL_LOCK_STATUS_CALIBRATING - fix naming scheme od notification interface functions - separate notifications for pins - rename attribute enum name: dplla -> dpll_a - rename pin-idx to pin-id - remove attributes: pin-parent-idx, device - replace bus-name and dev-name attributes with module-name - replace pin-label with 3 new attributes: pin-board-label, pin-panel-label, pin-package-label - add device-id-get and pin-id-get commands - remove rclk-dev-name atribute - rename DPLL_PIN_DIRECTION_SOURCE -> DPLL_PIN_DIRECTION_INPUT [03/10] dpll: core: Add DPLL framework base functions [04/10] dpll: netlink: Add DPLL framework base functions [05/10] dpll: api header: Add DPLL framework base - remove unspec attributes after removing from dpll netlink spec - move pin-direction to pin-dpll tuple - pass parent_priv on state_on_pin_<get/set> - align with new notification definitions from netlink spec - use separated notifications for dpll pins and devices - format notification messages as corresponding get netlink commands - rename pin-idx to pin-id - remove attributes pin-parent-idx, device - use DPLL_A_PIN_PARENT to hold information on parent pin or dpll device - refactor lookup for pins and dplls for dpll subsystem - replace bus-name, dev-name with module-name - replace pin-label with 3 new attributes: pin-board-label, pin-panel-label, pin-package-label - add device-id-get and pin-id-get commands - rename dpll_xa_lock to dpll_lock - improve doxygen in dpll_core.c - remove unused parent and dev fields from dpll_device struct - use u32 for pin_idx in dpll_pin_alloc - use driver provided pin properties struct - verify pin/dpll owner on registering pin - remove const arg modifier for helper _priv functions - remove function declaration _get_by_name() - update SPDX headers - parse netlink set attributes with nlattr array - remove rclk-dev-name attribute - remove device pointer from dpll_pin_register/dpll_device_register - remove redundant doxygen from dpll header - use module_name() to get name of module - add missing/remove outdated kdocs - fix call frequency_set only if available - fix call direction_set only for pin-dpll tuple [06/10] netdev: expose DPLL pin handle for netdevice - rebased on top of v8 changes - use dpll_msg_add_pin_handle() in dpll_pin_find_from_nlattr() and dpll_msg_add_pin_parents() - fixed handle to use DPLL_A_PIN_ID and removed temporary comments - added documentation record for dpll_pin pointer - fixed compilation of net/core/dev.c when CONFIG_DPLL is not enabled - adjusted patch description a bit [07/10] ice: add admin commands to access cgu configuration - Remove unspec attributes after removing from dpll netlink spec. [08/10] ice: implement dpll interface to control cgu - remove unspec attributes - do not store pin flags received in set commands - use pin state field to provide pin state to the caller - remove include of uapi header - remove redundant check against null arguments - propagate lock function return value to the caller - use switch case instead of if statements - fix dev_dbg to dev_err for error cases - fix dpll/pin lookup on dpll subsytem callbacks - fix extack of dpll subsystem callbacks - remove double negation and variable cast - simplify ice_dpll_pin_state_set function - pass parent_priv on state_on_pin_<get/set> - remove parent hw_idx lookup - fix use const qualifier for dpll/dpll_pin ops - fix IS_ERR macros usage in ice_dpll - add notify previous source state change - fix mutex locking on releasing pins - use '|=' instead of '+=' when modifing capabilities field - rename ice_dpll_register_pins function - clock_id function to return clock ID on the stack instead of using an output variable - DPLL_LOCK_STATUS_CALIBRATING was removed, return: DPLL_LOCK_STATUS_LOCKED - if dpll was locked DPLL_LOCK_STATUS_LOCKED_HO_ACQ - if dpll was locked and holdover is acquired - propagate and use dpll_priv to obtain pf pointer in corresponding functions. - remove null check for pf pointer - adapt to `dpll: core: fix notification scheme` - expose pf related pin to corresponding netdevice - fix dpll init error path - fix dpll pins naming scheme `source` -> `input` - replace pin-label with pin-board-label - dpll remove parent and dev fields from dpll_device - remove device pointer from dpll_pin_register/dpll_device_register - rename DPLL_PIN_DIRECTION_SOURCE -> DPLL_PIN_DIRECTION_INPUT [09/10] ptp_ocp: implement DPLL ops - replace pin-label with pin-board-label - dpll remove parent and dev fields from dpll_device - remove device pointer from dpll_pin_register/dpll_device_register - rename DPLL_PIN_DIRECTION_SOURCE -> DPLL_PIN_DIRECTION_INPUT [10/10] mlx5: Implement SyncE support using DPLL infrastructure - rebased on top of v8 changes: - changed notification scheme - no need to fill pin label - implemented locked_ho_acq status - rename DPLL_PIN_DIRECTION_SOURCE -> DPLL_PIN_DIRECTION_INPUT - remove device pointer from dpll_pin_register/dpll_device_register - fixed MSEES register writes - adjusted pin state and lock state values reported - fixed a white space issue v6 -> v7: * YAML spec: - remove nested 'pin' attribute - clean up definitions on top of the latest changes * pin object: - pin xarray uses id provided by the driver - remove usage of PIN_IDX_INVALID in set function - source_pin_get() returns object instead of idx - fixes in frequency support API * device and pin operations are const now * small fixes in naming in Makefile and in the functions * single mutex for the subsystem to avoid possible ABBA locks * no special *_priv() helpers anymore, private data is passed as void* * no netlink filters by name anymore, only index is supported * update ptp_ocp and ice drivers to follow new API version * add mlx5e driver as a new customer of the subsystem v5 -> v6: * rework pin part to better fit shared pins use cases * add YAML spec to easy generate user-space apps * simple implementation in ptp_ocp is back again v4 -> v5: * fix code issues found during last reviews: - replace cookie with clock id - follow one naming schema in dpll subsys - move function comments to dpll_core.c, fix exports - remove single-use helper functions - merge device register with alloc - lock and unlock mutex on dpll device release - move dpll_type to uapi header - rename DPLLA_DUMP_FILTER to DPLLA_FILTER - rename dpll_pin_state to dpll_pin_mode - rename DPLL_MODE_FORCED to DPLL_MODE_MANUAL - remove DPLL_CHANGE_PIN_TYPE enum value * rewrite framework once again (Arkadiusz) - add clock class: Provide userspace with clock class value of DPLL with dpll device dump netlink request. Clock class is assigned by driver allocating a dpll device. Clock class values are defined as specified in: ITU-T G.8273.2/Y.1368.2 recommendation. - dpll device naming schema use new pattern: "dpll_%s_%d_%d", where: - %s - dev_name(parent) of parent device, - %d (1) - enum value of dpll type, - %d (2) - device index provided by parent device. - new muxed/shared pin registration: Let the kernel module to register a shared or muxed pin without finding it or its parent. Instead use a parent/shared pin description to find correct pin internally in dpll_core, simplifing a dpll API * Implement complex DPLL design in ice driver (Arkadiusz) * Remove ptp_ocp driver from the series for now v3 -> v4: * redesign framework to make pins dynamically allocated (Arkadiusz) * implement shared pins (Arkadiusz) v2 -> v3: * implement source select mode (Arkadiusz) * add documentation * implementation improvements (Jakub) v1 -> v2: * implement returning supported input/output types * ptp_ocp: follow suggestions from Jonathan * add linux-clk mailing list v0 -> v1: * fix code style and errors * add linux-arm mailing list ==================== Signed-off-by: -
Jiri Pirko authored
Implement SyncE support using newly introduced DPLL support. Make sure that each PFs/VFs/SFs probed with appropriate capability will spawn a dpll auxiliary device and register appropriate dpll device and pin instances. Signed-off-by:
Arkadiusz Kubalewski <arkadiusz.kubalewski@intel.com> Signed-off-by:
Vadim Fedorenko <vadim.fedorenko@linux.dev> Signed-off-by:
-
Vadim Fedorenko authored
Implement basic DPLL operations in ptp_ocp driver as the simplest example of using new subsystem. Signed-off-by:
-
Arkadiusz Kubalewski authored
Control over clock generation unit is required for further development of Synchronous Ethernet feature. Interface provides ability to obtain current state of a dpll, its sources and outputs which are pins, and allows their configuration. Co-developed-by:
Milena Olech <milena.olech@intel.com> Signed-off-by:
Michal Michalik <michal.michalik@intel.com> Signed-off-by:
-
Arkadiusz Kubalewski authored
Add firmware admin command to access clock generation unit configuration, it is required to enable Extended PTP and SyncE features in the driver. Add definitions of possible hardware variations of input and output pins related to clock generation unit and functions to access the data. Signed-off-by:
-
Jiri Pirko authored
In case netdevice represents a SyncE port, the user needs to understand the connection between netdevice and associated DPLL pin. There might me multiple netdevices pointing to the same pin, in case of VF/SF implementation. Add a IFLA Netlink attribute to nest the DPLL pin handle, similar to how it is implemented for devlink port. Add a struct dpll_pin pointer to netdev and protect access to it by RTNL. Expose netdev_dpll_pin_set() and netdev_dpll_pin_clear() helpers to the drivers so they can set/clear the DPLL pin relationship to netdev. Note that during the lifetime of struct dpll_pin the pin handle does not change. Therefore it is save to access it lockless. It is drivers responsibility to call netdev_dpll_pin_clear() before dpll_pin_put(). Signed-off-by:
-
Vadim Fedorenko authored
DPLL framework is used to represent and configure DPLL devices in systems. Each device that has DPLL and can configure inputs and outputs can use this framework. Implement dpll netlink framework functions for enablement of dpll subsystem netlink family. Co-developed-by:
-
Vadim Fedorenko authored
DPLL framework is used to represent and configure DPLL devices in systems. Each device that has DPLL and can configure inputs and outputs can use this framework. Implement core framework functions for further interactions with device drivers implementing dpll subsystem, as well as for interactions of DPLL netlink framework part with the subsystem itself. Co-developed-by:
-
Vadim Fedorenko authored
Add a protocol spec for DPLL. Add code generated from the spec. Signed-off-by:
Jakub Kicinski <kuba@kernel.org> Signed-off-by:
-
Vadim Fedorenko authored
Add documentation explaining common netlink interface to configure DPLL devices and monitoring events. Common way to implement DPLL device in a driver is also covered. Co-developed-by:
Bagas Sanjaya <bagasdotme@gmail.com> Signed-off-by:
-
git://git.kernel.org/pub/scm/linux/kernel/git/tnguy/nextDavid S. Miller authored
-queue Tony Nguyen says: ==================== Support rx-fcs on/off for VFs Ahmed Zaki says: Allow the user to turn on/off the CRC/FCS stripping through ethtool. We first add the CRC offload capability in the virtchannel, then the feature is enabled in ice and iavf drivers. We make sure that the netdev features are fixed such that CRC stripping cannot be disabled if VLAN rx offload (VLAN strip) is enabled. Also, VLAN stripping cannot be enabled unless CRC stripping is ON. Testing was done using tcpdump to make sure that the CRC is included in the frame after: # ethtool -K <interface> rx-fcs on and is not included when it is back "off". Also, ethtool should return an error for the above command if "rx-vlan-offload" is already on and at least one VLAN interface/filter exists on the VF. ==================== Signed-off-by:
-
- 16 Sep, 2023 15 commits
-
-
David S. Miller authored
Aananth V says: ==================== tcp: new TCP_INFO stats for RTO events The 2023 SIGCOMM paper "Improving Network Availability with Protective ReRoute" has indicated Linux TCP's RTO-triggered txhash rehashing can effectively reduce application disruption during outages. To better measure the efficacy of this feature, this patch set adds three more detailed stats during RTO recovery and exports via TCP_INFO. Applications and monitoring systems can leverage this data to measure the network path diversity and end-to-end repair latency during network outages to improve their network infrastructure. Patch 1 fixes a bug in TFO SYNACK that we encountered while testing these new metrics. Patch 2 adds the new metrics to tcp_sock and tcp_info. v2: Addressed feedback from a check bot in patch 2 by removing the inline keyword from the tcp_update_rto_time and tcp_update_rto_stats functions. Changed a comment in include/net/tcp.h to fit under 80 words. ==================== Signed-off-by: -
Aananth V authored
The 2023 SIGCOMM paper "Improving Network Availability with Protective ReRoute" has indicated Linux TCP's RTO-triggered txhash rehashing can effectively reduce application disruption during outages. To better measure the efficacy of this feature, this patch adds three more detailed stats during RTO recovery and exports via TCP_INFO. Applications and monitoring systems can leverage this data to measure the network path diversity and end-to-end repair latency during network outages to improve their network infrastructure. The following counters are added to tcp_sock in order to track RTO events over the lifetime of a TCP socket. 1. u16 total_rto - Counts the total number of RTO timeouts. 2. u16 total_rto_recoveries - Counts the total number of RTO recoveries. 3. u32 total_rto_time - Counts the total time spent (ms) in RTO recoveries. (time spent in CA_Loss and CA_Recovery states) To compute total_rto_time, we add a new u32 rto_stamp field to tcp_sock. rto_stamp records the start timestamp (ms) of the last RTO recovery (CA_Loss). Corresponding fields are also added to the tcp_info struct. Signed-off-by:Aananth V <aananthv@google.com> Signed-off-by:
Neal Cardwell <ncardwell@google.com> Signed-off-by:
Yuchung Cheng <ycheng@google.com> Reviewed-by:
Eric Dumazet <edumazet@google.com> Signed-off-by:
-
Aananth V authored
For passive TCP Fast Open sockets that had SYN/ACK timeout and did not send more data in SYN_RECV, upon receiving the final ACK in 3WHS, the congestion state may awkwardly stay in CA_Loss mode unless the CA state was undone due to TCP timestamp checks. However, if tcp_rcv_synrecv_state_fastopen() decides not to undo, then we should enter CA_Open, because at that point we have received an ACK covering the retransmitted SYNACKs. Currently, the icsk_ca_state is only set to CA_Open after we receive an ACK for a data-packet. This is because tcp_ack does not call tcp_fastretrans_alert (and tcp_process_loss) if !prior_packets Note that tcp_process_loss() calls tcp_try_undo_recovery(), so having tcp_rcv_synrecv_state_fastopen() decide that if we're in CA_Loss we should call tcp_try_undo_recovery() is consistent with that, and low risk. Fixes: dad8cea7 ("tcp: fix TFO SYNACK undo to avoid double-timestamp-undo") Signed-off-by:
-
David S. Miller authored
Oleksij Rempel says: ==================== net: dsa: microchip: add drive strength support changes v5: - rename milliamp to microamp - do not expect negative error code on snprintf - set coma after last struct element - rename found to have_any_prop changes v4: - integrate microchip feedback to the ksz9477_drive_strengths comment. - add Reviewed-by: Rob Herring <robh@kernel.org> changes v3: - yaml: use enum instead of min/max - do not use snprintf() on overlapping buffer. - unify ksz_drive_strength_to_reg() and ksz_drive_strength_error(). Make it usable for KSZ9477 and KSZ8830 variants. - use ksz_rmw8() in ksz9477_drive_strength_write() ==================== Signed-off-by: -
Oleksij Rempel authored
Add device tree based drive strength configuration support. It is needed to pass EMI validation on our hardware. Configuration values are based on the vendor's reference driver. Tested on KSZ9563R. Signed-off-by:
Oleksij Rempel <o.rempel@pengutronix.de> Reviewed-by:
Andrew Lunn <andrew@lunn.ch> Reviewed-by:
Vladimir Oltean <olteanv@gmail.com> Signed-off-by:
-
Oleksij Rempel authored
Extend device tree bindings to support drive strength configuration for the ksz* switches. Introduced properties: - microchip,hi-drive-strength-microamp: Controls the drive strength for high-speed interfaces like GMII/RGMII and more. - microchip,lo-drive-strength-microamp: Governs the drive strength for low-speed interfaces such as LEDs, PME_N, and others. - microchip,io-drive-strength-microamp: Controls the drive strength for for undocumented Pads on KSZ88xx variants. Signed-off-by:
Rob Herring <robh@kernel.org> Reviewed-by:
-
git://git.kernel.org/pub/scm/linux/kernel/git/tnguy/next-queueDavid S. Miller authored
Tony Nguyen says: ==================== Introduce Intel IDPF driver Pavan Kumar Linga says: This patch series introduces the Intel Infrastructure Data Path Function (IDPF) driver. It is used for both physical and virtual functions. Except for some of the device operations the rest of the functionality is the same for both PF and VF. IDPF uses virtchnl version2 opcodes and structures defined in the virtchnl2 header file which helps the driver to learn the capabilities and register offsets from the device Control Plane (CP) instead of assuming the default values. The format of the series follows the driver init flow to interface open. To start with, probe gets called and kicks off the driver initialization by spawning the 'vc_event_task' work queue which in turn calls the 'hard reset' function. As part of that, the mailbox is initialized which is used to send/receive the virtchnl messages to/from the CP. Once that is done, 'core init' kicks in which requests all the required global resources from the CP and spawns the 'init_task' work queue to create the vports. Based on the capability information received, the driver creates the said number of vports (one or many) where each vport is associated to a netdev. Also, each vport has its own resources such as queues, vectors etc. From there, rest of the netdev_ops and data path are added. IDPF implements both single queue which is traditional queueing model as well as split queue model. In split queue model, it uses separate queue for both completion descriptors and buffers which helps to implement out-of-order completions. It also helps to implement asymmetric queues, for example multiple RX completion queues can be processed by a single RX buffer queue and multiple TX buffer queues can be processed by a single TX completion queue. In single queue model, same queue is used for both descriptor completions as well as buffer completions. It also supports features such as generic checksum offload, generic receive offload (hardware GRO) etc. --- v7: Patch 2: * removed pci_[disable|enable]_pcie_error_reporting as they are dropped from the core Patch 4, 9: * used 'kasprintf' instead of 'snprintf' to avoid providing explicit character string size which also fixes "-Wformat-truncation" warnings Patch 14: * used 'ethtool_sprintf' instead of 'snprintf' to avoid providing explicit character string size which also fixes "-Wformat-truncation" warning * add string format argument to the 'ethtool_sprintf' to avoid warning on "-Wformat-security" v6: https://lore.kernel.org/netdev/20230825235954.894050-1-pavan.kumar.linga@intel.com/ Note: 'Acked-by' was only added to patches 1, 2, 12 and not to the other patches because of the changes in v6 Patch 3, 4, 5, 6, 7, 8, 9, 11, 13, 14, 15: * renamed 'reset_lock' to 'vport_ctrl_lock' to reflect the lock usage * to avoid defensive programming, used 'vport_ctrl_lock' for the user callbacks that access the 'vport' to prevent the hardware reset thread from releasing the 'vport', when the user callback is in progress * added some variables to netdev private structure to avoid vport access if possible from ethtool and ndo callbacks * moved 'mac_filter_list_lock' and MAC related flags to vport_config structure and refactored mac filter flow to handle asynchronous ndo mac filter callbacks * stop the queues before starting the reset flow to avoid TX hangs * removed 'sw_mutex' and 'stop_mutex' as they are not needed anymore * added missing clear bit in 'init_task' error path * renamed labels appropriately Patch 8: * replaced page_pool_put_page with page_pool_put_full_page * for the page pool max_len, used PAGE_SIZE Patch 10, 11, 13: * made use of the 'netif_txq_maybe_stop', '__netif_txq_completed_wake' helper macros Patch 13: * removed IDPF_HR_RESET_IN_PROG flag check in idpf_tx_singleq_start as it is defensive Patch 14: * removed max descriptor check as the core does that * removed unnecessary error messages * removed the stats that are common between the ones reported by ethtool and ip link * replaced snprintf with ethtool_sprintf * added a comment to explain the reason for the max queue check * as the netdev queues are set on alloc, there is no need to set them again on reset unless there is a queue change, so move the 'idpf_set_real_num_queues' to 'idpf_initiate_soft_reset' Patch 15: * reworded the 'configure SRIOV' in the commit message v5: https://lore.kernel.org/netdev/20230816004305.216136-1-anthony.l.nguyen@intel.com/ Most Patches: * wrapped line limit to 80 chars to those which don't effect readability Patch 12: * in skb_add_rx_frag, offset 'headlen' w.r.t page_offset when adding a frag to avoid adding the header again Patch 14: * added NULL check for 'rxq' when dereferencing it in page_pool_get_stats v4: https://lore.kernel.org/netdev/20230808003416.3805142-1-anthony.l.nguyen@intel.com/ Patch 1: * s/virtcnl/virtchnl * removed the kernel doc for the error code definitions that don't exist * reworded the summary part in the virtchnl2 header Patch 3: * don't set local variable to NULL on error * renamed sq_send_command_out label with err_unlock * don't use __GFP_ZERO in dma_alloc_coherent Patch 4: * introduced mailbox workqueue to process mailbox interrupts Patch 3, 4, 5, 6, 7, 8, 9, 11, 15: * removed unnecessary variable 0-init Patch 3, 5, 7, 8, 9, 15: * removed defensive programming checks wherever applicable * removed IDPF_CAP_FIELD_LAST as it can be treated as defensive programming Patch 3, 4, 5, 6, 7: * replaced IDPF_DFLT_MBX_BUF_SIZE with IDPF_CTLQ_MAX_BUF_LEN Patch 2 to 15: * add kernel-doc for idpf.h and idpf_txrx.h enums and structures Patch 4, 5, 15: * adjusted the destroy sequence of the workqueues as per the alloc sequence Patch 4, 5, 9, 15: * scrub unnecessary flags in 'idpf_flags' - IDPF_REMOVE_IN_PROG flag can take care of the cases where IDPF_REL_RES_IN_PROG is used, removed the later one - IDPF_REQ_[TX|RX]_SPLITQ are replaced with struct variables - IDPF_CANCEL_[SERVICE|STATS]_TASK are redundant as the work queue doesn't get rescheduled again after 'cancel_delayed_work_sync' - IDPF_HR_CORE_RESET is removed as there is no set_bit for this flag - IDPF_MB_INTR_TRIGGER is removed as it is not needed anymore with the mailbox workqueue implementation Patch 7 to 15: * replaced the custom buffer recycling code with page pool API * switched the header split buffer allocations from using a bunch of pages to using one large chunk of DMA memory * reordered some of the flows in vport_open to support page pool Patch 8, 12: * don't suppress the alloc errors by using __GFP_NOWARN Patch 9: * removed dyn_ctl_clrpba_m as it is not being used Patch 14: * introduced enum idpf_vport_reset_cause instead of using vport flags * introduced page pool stats v3: https://lore.kernel.org/netdev/20230616231341.2885622-1-anthony.l.nguyen@intel.com/ Patch 5: * instead of void, used 'struct virtchnl2_create_vport' type for vport_params_recvd and vport_params_reqd and removed the typecasting * used u16/u32 as needed instead of int for variables which cannot be negative and updated in all the places whereever applicable Patch 6: * changed the commit message to "add ptypes and MAC filter support" * used the sender Signed-off-by as the last tag on all the patches * removed unnecessary variables 0-init * instead of fixing the code in this commit, fixed it in the commit where the change was introduced first * moved get_type_info struct on to the stack instead of memory alloc * moved mutex_lock and ptype_info memory alloc outside while loop and adjusted the return flow * used 'break' instead of 'continue' in ptype id switch case v2: https://lore.kernel.org/netdev/20230614171428.1504179-1-anthony.l.nguyen@intel.com/ Patch 2: * added "Intel(R)" to the DRV_SUMMARY and Makefile. Patch 4, 5, 6, 15: * replaced IDPF_VC_MSG_PENDING flag with mutex 'vc_buf_lock' for the adapter related virtchnl opcodes. * get the mutex lock in the virtchnl send thread itself instead of in receive thread. Patch 5, 6, 7, 8, 9, 11, 14, 15: * replaced IDPF_VPORT_VC_MSG_PENDING flag with mutex 'vc_buf_lock' for the vport related virtchnl opcodes. * get the mutex lock in the virtchnl send thread itself instead of in receive thread. Patch 6: * converted get_ptype_info logic from 1:N to 1:1 message exchange for better handling of mutex lock. Patch 15: * introduced 'stats_lock' spinlock to avoid concurrent stats update. v1: https://lore.kernel.org/netdev/20230530234501.2680230-1-anthony.l.nguyen@intel.com/ ==================== Signed-off-by:
-
David S. Miller authored
Keguang Zhang says: ==================== Move Loongson1 MAC arch-code to the driver dir In order to convert Loongson1 MAC platform devices to the devicetree nodes, Loongson1 MAC arch-code should be moved to the driver dir. Add dt-binding document and update MAINTAINERS file accordingly. In other words, this patchset is a preparation for converting Loongson1 platform devices to devicetree. Changelog V4 -> V5: Replace stmmac_probe_config_dt() with devm_stmmac_probe_config_dt() Replace stmmac_pltfr_probe() with devm_stmmac_pltfr_probe() Squash patch 4 into patch 2 and 3 V3 -> V4: Add Acked-by tag from Krzysztof Kozlowski Add "|" to description part Amend "phy-mode" property Drop ls1x_dwmac_syscon definition and its instances Drop three redundant fields from the ls1x_dwmac structure Drop the ls1x_dwmac_init() method. Update the dt-binding document entry of Loongson1 Ethernet Some minor improvements V2 -> V3: Split the DT-schema file into loongson,ls1b-gmac.yaml and loongson,ls1c-emac.yaml (suggested by Serge Semin) Change the compatibles to loongson,ls1b-gmac and loongson,ls1c-emac Rename loongson,dwmac-syscon to loongson,ls1-syscon Amend the title Add description Add Reviewed-by tag from Krzysztof Kozlowski Change compatibles back to loongson,ls1b-syscon and loongson,ls1c-syscon Determine the device ID by physical base address(suggested by Serge Semin) Use regmap instead of regmap fields Use syscon_regmap_lookup_by_phandle() Some minor fixes Update the entries of MAINTAINERS V1 -> V2: Leave the Ethernet platform data for now Make the syscon compatibles more specific Fix "clock-names" and "interrupt-names" property Rename the syscon property to "loongson,dwmac-syscon" Drop "phy-handle" and "phy-mode" requirement Revert adding loongson,ls1b-dwmac/loongson,ls1c-dwmac to snps,dwmac.yaml Fix the build errors due to CONFIG_OF being unset Change struct reg_field definitions to const Rename the syscon property to "loongson,dwmac-syscon" Add MII PHY mode for LS1C Improve the commit message ==================== Signed-off-by: -
Keguang Zhang authored
This glue driver is created based on the arch-code implemented earlier with the platform-specific settings. Use syscon for SYSCON register access. And modify MAINTAINERS to add a new F: entry for this driver. Partially based on the previous work by Serge Semin. Signed-off-by:
Keguang Zhang <keguang.zhang@gmail.com> Reviewed-by:
Serge Semin <fancer.lancer@gmail.com> Signed-off-by:
-
Keguang Zhang authored
Add devicetree binding document for Loongson-1 Ethernet controller. And modify MAINTAINERS to add a new F: entry for Loongson1 dt-binding documents. Signed-off-by:
Krzysztof Kozlowski <krzysztof.kozlowski@linaro.org> Reviewed-by:
-
Keguang Zhang authored
Add Loongson LS1B and LS1C compatibles for system controller. Signed-off-by:
-
Alex Austin authored
Follow the style used in the core kernel (e.g. include/linux/etherdevice.h and include/linux/in6.h) for the PTP IPv6 and Ethernet addresses. No functional changes. Signed-off-by:
Alex Austin <alex.austin@amd.com> Reviewed-by:
Edward Cree <ecree.xilinx@gmail.com> Signed-off-by:
-
Lorenzo Bianconi authored
Check if wlan.offload_enable and wlan.offload_disable callbacks are set in mtk_wed_flow_add/mtk_wed_flow_remove since mt7996 will not rely on them. Signed-off-by:
Lorenzo Bianconi <lorenzo@kernel.org> Reviewed-by:
Simon Horman <horms@kernel.org> Signed-off-by:
-
Eric Dumazet authored
It can be time consuming to track driver bugs, that might be detected too late from this confusing warning in skb_try_coalesce() WARN_ON_ONCE(delta < len); Add sanity check in skb_add_rx_frag() and skb_coalesce_rx_frag() to better track bug origin for CONFIG_DEBUG_NET=y builds. Signed-off-by:
-
Eric Dumazet authored
When adding sk->sk_prot->release_cb() call from __sk_flush_backlog() Paolo suggested using indirect call helpers to take care of CONFIG_RETPOLINE=y case. It turns out Google had such mitigation for years in release_sock(), it is time to make this public :) Suggested-by:
Paolo Abeni <pabeni@redhat.com> Signed-off-by:
-
- 15 Sep, 2023 2 commits
-
-
Dan Carpenter authored
Indent this if statement one tab. Signed-off-by:
Dan Carpenter <dan.carpenter@linaro.org> Signed-off-by:
-
David S. Miller authored
==================== net: Add Half Duplex support for ICSSG Driver This series adds support for half duplex operation for ICSSG driver. In order to support half-duplex operation at 10M and 100M link speeds, the PHY collision detection signal (COL) should be routed to ICSSG GPIO pin (PRGx_PRU0/1_GPI10) so that firmware can detect collision signal and apply the CSMA/CD algorithm applicable for half duplex operation. A DT property, "ti,half-duplex-capable" is introduced for this purpose in the first patch of the series. If board has PHY COL pin conencted to PRGx_PRU1_GPIO10, this DT property can be added to eth node of ICSSG, MII port to support half duplex operation at that port. Second patch of the series configures driver to support half-duplex operation if the DT property "ti,half-duplex-capable" is enabled. This series addresses comments on [v2]. This series is based on the latest net-next/main. This series has no dependency. Changes from v1 to v2: *) Changed the description of "ti,half-duplex-capable" property as asked by Rob and Andrew to avoid confusion between capable and enable. Changes from v1 to v2: *) Dropped the RFC tag. *) Added RB tags of Andrew and Roger. [v1] https://lore.kernel.org/all/20230830113134.1226970-1-danishanwar@ti.com/ [v2] https://lore.kernel.org/all/20230911060200.2164771-1-danishanwar@ti.com/ ==================== Signed-off-by:
-
