Changes in Makefile

2a82874a - ice: add Tx hang devlink health reporter

Mon, 16 Dec 2024 14:15:34 +0000

ice: add Tx hang devlink health reporterAdd Tx hang devlink health reporter, see struct ice_tx_hang_event to seewhat exactly is reported. For now dump descriptors with little metadataand skb diagnostic information.Reviewed-by: Igor Bagnucki Reviewed-by: Wojciech Drewek Co-developed-by: Mateusz Polchlopek Signed-off-by: Mateusz Polchlopek Tested-by: Pucha Himasekhar Reddy (A Contingent worker at Intel)Signed-off-by: Przemek Kitszel Signed-off-by: Tony Nguyen List of files: /linux-6.15/drivers/net/ethernet/intel/ice/Makefile

2846fe56 - ice: rename devlink_port.[ch] to port.[ch]

Mon, 16 Dec 2024 14:15:33 +0000

ice: rename devlink_port.[ch] to port.[ch]Drop "devlink_" prefix from files that sit in devlink/.I'm going to add more files there, and repeating "devlink" does not feelgood. This is also the scheme used in most other places, most notably thedevlink core files are named like that.devlink.[ch] stays as is.Reviewed-by: Kalesh AP Signed-off-by: Przemek Kitszel Signed-off-by: Tony Nguyen List of files: /linux-6.15/drivers/net/ethernet/intel/ice/Makefile

0c6a3cb6 - ice: basic support for VLAN in subfunctions

Tue, 20 Aug 2024 06:57:59 +0000

ice: basic support for VLAN in subfunctionsImplement add / delete vlan for subfunction type VSI.Reviewed-by: Simon Horman Reviewed-by: Wojciech Drewek Signed-off-by: Michal Swiatkowski Tested-by: Rafal Romanowski Signed-off-by: Tony Nguyen List of files: /linux-6.15/drivers/net/ethernet/intel/ice/Makefile

177ef7f1 - ice: base subfunction aux driver

Tue, 20 Aug 2024 06:57:51 +0000

ice: base subfunction aux driverImplement subfunction driver. It is probe when subfunction port isactivated.VSI is already created. During the probe VSI is being configured.MAC unicast and broadcast filter is added to allow traffic to pass.Store subfunction pointer in VSI struct. The same is done for VFpointer. Make union of subfunction and VF pointer as only one of themcan be set with one VSI.Reviewed-by: Simon Horman Reviewed-by: Jiri Pirko Signed-off-by: Piotr Raczynski Signed-off-by: Michal Swiatkowski Tested-by: Rafal Romanowski Signed-off-by: Tony Nguyen List of files: /linux-6.15/drivers/net/ethernet/intel/ice/Makefile

9a4c07aa - ice: add parser execution main loop

Thu, 25 Jul 2024 22:08:01 +0000

ice: add parser execution main loopImplement the core work of the runtime parser via:- ice_parser_rt_execute()- ice_parser_rt_reset()- ice_parser_rt_pkt_buf_set()Reviewed-by: Marcin Szycik Signed-off-by: Qi Zhang Signed-off-by: Junfeng Guo Signed-off-by: Ahmed Zaki Tested-by: Rafal Romanowski Reviewed-by: Simon Horman Signed-off-by: Tony Nguyen List of files: /linux-6.15/drivers/net/ethernet/intel/ice/Makefile

86ff3d79 - ice: add parser create and destroy skeleton

Thu, 25 Jul 2024 22:07:57 +0000

ice: add parser create and destroy skeletonAdd new parser module which can parse a packet in binary and generateinformation like ptype, protocol/offset pairs and flags which can be laterused to feed the FXP profile creation directly.Add skeleton of the create and destroy APIs:ice_parser_create()ice_parser_destroy()Reviewed-by: Simon Horman Reviewed-by: Marcin Szycik Reviewed-by: Wojciech Drewek Signed-off-by: Qi Zhang Signed-off-by: Junfeng Guo Co-developed-by: Ahmed Zaki Signed-off-by: Ahmed Zaki Tested-by: Rafal Romanowski Signed-off-by: Tony Nguyen List of files: /linux-6.15/drivers/net/ethernet/intel/ice/Makefile

4ebc5f25 - ice: move devlink port code to a separate file

Mon, 25 Mar 2024 21:34:32 +0000

ice: move devlink port code to a separate fileKeep devlink related code in a separate file. More devlink port code andhandlers will be added here for other port operations.Remove no longer needed include of our devlink.h in ice_lib.c.Reviewed-by: Przemek Kitszel Reviewed-by: Wojciech Drewek Signed-off-by: Piotr Raczynski Signed-off-by: Michal Swiatkowski Reviewed-by: Simon Horman Tested-by: Pucha Himasekhar Reddy (A Contingent worker at Intel)Signed-off-by: Tony Nguyen List of files: /linux-6.15/drivers/net/ethernet/intel/ice/Makefile

0545cc86 - ice: move ice_devlink.[ch] to devlink folder

Mon, 25 Mar 2024 21:34:31 +0000

ice: move ice_devlink.[ch] to devlink folderOnly moving whole files, fixing Makefile and bunch of includes.Some changes to ice_devlink file was done even in representor part (Txtopology), so keep it as final patch to not mess up with rebasing.After moving to devlink folder there is no need to have such long namefor these files. Rename them to simple devlink.Reviewed-by: Aleksandr Loktionov Signed-off-by: Michal Swiatkowski Reviewed-by: Simon Horman Tested-by: Pucha Himasekhar Reddy (A Contingent worker at Intel)Signed-off-by: Tony Nguyen List of files: /linux-6.15/drivers/net/ethernet/intel/ice/Makefile

0e2bddf9 - ice: add ice_adapter for shared data across PFs on the same NIC

Mon, 25 Mar 2024 23:20:37 +0000

ice: add ice_adapter for shared data across PFs on the same NICThere is a need for synchronization between ice PFs on the same physicaladapter.Add a "struct ice_adapter" for holding data shared between PFs of thesame multifunction PCI device. The struct is refcounted - each ice_pfholds a reference to it.Its first use will be for PTP. I expect it will be useful also toimprove the ugliness that is ice_prot_id_tbl.Reviewed-by: Przemek Kitszel Signed-off-by: Michal Schmidt Tested-by: Pucha Himasekhar Reddy (A Contingent worker at Intel)Signed-off-by: Tony Nguyen List of files: /linux-6.15/drivers/net/ethernet/intel/ice/Makefile

96a9a934 - ice: configure FW logging

Wed, 13 Dec 2023 05:07:12 +0000

ice: configure FW loggingUsers want the ability to debug FW issues by retrieving theFW logs from the E8xx devices. Use debugfs to allow the user toconfigure the log level and number of messages for FW logging.If FW logging is supported on the E8xx then the file 'fwlog' will becreated under the PCI device ID for the ice driver. If the file does notexist then either the E8xx doesn't support FW logging or debugfs is notenabled on the system.One thing users want to do is control which events are reported. Theuser can read and write the 'fwlog/modules/' to get/setthe log levels. Each module in the FW that supports logging ht as a fileunder 'fwlog/modules' that supports reading (to see what the current loglevel is) and writing (to change the log level).The format to set the log levels for a module are: # echo > /sys/kernel/debug/ice/0000\:18\:00.0/fwlog/modules/The supported log levels are: * none * error * warning * normal * verboseEach level includes the messages from the previous/lower levelThe modules that are supported are: * general * ctrl * link * link_topo * dnl * i2c * sdp * mdio * adminq * hdma * lldp * dcbx * dcb * xlr * nvm * auth * vpd * iosf * parser * sw * scheduler * txq * rsvd * post * watchdog * task_dispatch * mng * synce * health * tsdrv * pfreg * mdlver * allThe module 'all' is a special module which allows the user to read orwrite to all of the modules.The following example command would set the DCB module to the 'normal'log level: # echo normal > /sys/kernel/debug/ice/0000\:18\:00.0/fwlog/modules/dcbIf the user wants to set the DCB, Link, and the AdminQ modules to'verbose' then the commands are: # echo verbose > /sys/kernel/debug/ice/0000\:18\:00.0/fwlog/modules/dcb # echo verbose > /sys/kernel/debug/ice/0000\:18\:00.0/fwlog/modules/link # echo verbose > /sys/kernel/debug/ice/0000\:18\:00.0/fwlog/modules/adminqIf the user wants to set all modules to the 'warning' level then thecommand is: # echo warning > /sys/kernel/debug/ice/0000\:18\:00.0/fwlog/modules/allIf the user wants to disable logging for a module then they can set thelevel to 'none'. An example setting the 'watchdog' module is: # echo none > /sys/kernel/debug/ice/0000\:18\:00.0/fwlog/modules/watchdogIf the user wants to see what the log level is for a specific modulethen the command is: # cat /sys/kernel/debug/ice/0000\:18\:00.0/fwlog/modules/dcbThis will return the log level for the DCB module. If the user wants tosee the log level for all the modules then the command is: # cat /sys/kernel/debug/ice/0000\:18\:00.0/fwlog/modules/allWriting to the module file will update the configuration, but NOT enable theconfiguration (that is a separate command).In addition to configuring the modules, the user can also configure thenumber of log messages (nr_messages) to include in a single Admin ReceiveQueue (ARQ) event.The range is 1-128 (1 means push every log message, 128means push only when the max AQ command buffer is full). The suggestedvalue is 10.To see/change the resolution the user can read/write the'fwlog/nr_messages' file. An example changing the value to 50 is # echo 50 > /sys/kernel/debug/ice/0000\:18\:00.0/fwlog/nr_messagesTo see the current value of 'nr_messages' then the command is: # cat /sys/kernel/debug/ice/0000\:18\:00.0/fwlog/nr_messagesSigned-off-by: Paul M Stillwell Jr Tested-by: Pucha Himasekhar Reddy (A Contingent worker at Intel)Signed-off-by: Tony Nguyen List of files: /linux-6.15/drivers/net/ethernet/intel/ice/Makefile

4da71a77 - ice: read internal temperature sensor

Fri, 01 Dec 2023 18:08:39 +0000

ice: read internal temperature sensorSince 4.30 firmware exposes internal thermal sensor reading via adminqueue commands. Expose those readouts via hwmon API when supported.Datasheet:Get Sensor Reading Command (Opcode: 0x0632)+--------------------+--------+--------------------+-------------------------+| Name | Bytes | Value | Remarks |+--------------------+--------+--------------------+-------------------------+| Flags | 1-0 | | || Opcode | 2-3 | 0x0632 | Command opcode || Datalen | 4-5 | 0 | No external buffer. || Return value | 6-7 | | Return value. || Cookie High | 8-11 | Cookie | || Cookie Low | 12-15 | Cookie | || Sensor | 16 | | 0x00: Internal temp || | | | 0x01-0xFF: Reserved. || Format | 17 | Requested response | Only 0x00 is supported. || | | format | 0x01-0xFF: Reserved. || Reserved | 18-23 | | || Data Address high | 24-27 | Response buffer | || | | address | || Data Address low | 28-31 | Response buffer | || | | address | |+--------------------+--------+--------------------+-------------------------+Get Sensor Reading Response (Opcode: 0x0632)+--------------------+--------+--------------------+-------------------------+| Name | Bytes | Value | Remarks |+--------------------+--------+--------------------+-------------------------+| Flags | 1-0 | | || Opcode | 2-3 | 0x0632 | Command opcode || Datalen | 4-5 | 0 | No external buffer || Return value | 6-7 | | Return value. || | | | EINVAL: Invalid || | | | parameters || | | | ENOENT: Unsupported || | | | sensor || | | | EIO: Sensor access || | | | error || Cookie High | 8-11 | Cookie | || Cookie Low | 12-15 | Cookie | || Sensor Reading | 16-23 | | Format of the reading || | | | is dependent on request || Data Address high | 24-27 | Response buffer | || | | address | || Data Address low | 28-31 | Response buffer | || | | address | |+--------------------+--------+--------------------+-------------------------+Sensor Reading for Sensor 0x00 (Internal Chip Temperature):+--------------------+--------+--------------------+-------------------------+| Name | Bytes | Value | Remarks |+--------------------+--------+--------------------+-------------------------+| Thermal Sensor | 0 | | Reading in degrees || reading | | | Celsius. Signed int8 || Warning High | 1 | | Warning High threshold || threshold | | | in degrees Celsius. || | | | Unsigned int8. || | | | 0xFF when unsupported || Critical High | 2 | | Critical High threshold || threshold | | | in degrees Celsius. || | | | Unsigned int8. || | | | 0xFF when unsupported || Fatal High | 3 | | Fatal High threshold || threshold | | | in degrees Celsius. || | | | Unsigned int8. || | | | 0xFF when unsupported || Reserved | 4-7 | | |+--------------------+--------+--------------------+-------------------------+Driver provides current reading from HW as well as device specificthresholds for thermal alarm (Warning, Critical, Fatal) events.$ sensorsOutput=========================================================ice-pci-b100Adapter: PCI adaptertemp1: +62.0°C (high = +95.0°C, crit = +105.0°C) (emerg = +115.0°C)Tested on Intel Corporation Ethernet Controller E810-C for SFPCo-developed-by: Marcin Domagala Signed-off-by: Marcin Domagala Co-developed-by: Eric Joyner Signed-off-by: Eric Joyner Reviewed-by: Marcin Szycik Reviewed-by: Przemek Kitszel Signed-off-by: Konrad Knitter Tested-by: Pucha Himasekhar Reddy (A Contingent worker at Intel)Signed-off-by: Tony Nguyen Signed-off-by: Paolo Abeni List of files: /linux-6.15/drivers/net/ethernet/intel/ice/Makefile

91e43ca0 - ice: fix linking when CONFIG_PTP_1588_CLOCK=n

Mon, 02 Oct 2023 18:51:32 +0000

ice: fix linking when CONFIG_PTP_1588_CLOCK=nThe recent support for DPLL introduced by commit 8a3a565ff210 ("ice: addadmin commands to access cgu configuration") and commit d7999f5ea64b ("ice:implement dpll interface to control cgu") broke linking the ice driver ifCONFIG_PTP_1588_CLOCK=n:ld: vmlinux.o: in function `ice_init_feature_support':(.text+0x8702b8): undefined reference to `ice_is_phy_rclk_present'ld: (.text+0x8702cd): undefined reference to `ice_is_cgu_present'ld: (.text+0x8702d9): undefined reference to `ice_is_clock_mux_present_e810t'ld: vmlinux.o: in function `ice_dpll_init_info_direct_pins':ice_dpll.c:(.text+0x894167): undefined reference to `ice_cgu_get_pin_freq_supp'ld: ice_dpll.c:(.text+0x894197): undefined reference to `ice_cgu_get_pin_name'ld: ice_dpll.c:(.text+0x8941a8): undefined reference to `ice_cgu_get_pin_type'ld: vmlinux.o: in function `ice_dpll_update_state':ice_dpll.c:(.text+0x894494): undefined reference to `ice_get_cgu_state'ld: vmlinux.o: in function `ice_dpll_init':(.text+0x8953d5): undefined reference to `ice_get_cgu_rclk_pin_info'The first commit broke things by calling functions inice_init_feature_support that are compiled as part of ice_ptp_hw.o,including:* ice_is_phy_rclk_present* ice_is_clock_mux_present_e810t* ice_is_cgU_presentThe second commit continued the break by calling several CGU functionsdefined in ice_ptp_hw.c in the DPLL code.Because the ice_dpll.c file is compiled unconditionally, it will notlink when CONFIG_PTP_1588_CLOCK=n.It might be possible to break this dependency and expose those functionswithout CONFIG_PTP_1588_CLOCK, but that is not clear to me.For the DPLL case, simply compile ice_dpll.o only when we haveCONFIG_PTP_1588_CLOCK. Add stub no-op implementation of ice_dpll_init() andice_dpll_uninit() when CONFIG_PTP_1588_CLOCK=n into ice_dpll.hThe other functions are part of checking the netlist to see if hardwarefeatures are enabled. These checks don't really belong in ice_ptp_hw.c, andmake more sense as part of the ice_common.c file. We already haveice_is_gps_in_netlist() in ice_common.c which is doing a similar check.Move the functions into ice_common.c and rename them to have the similarpostfix of "in_netlist()" to be more expressive of what they are actuallychecking.This also makes the ice_find_netlist_node only called from withinice_common.c, so its safe to mark it static and stop declaring it in theice_common.h header as well.Fixes: 8a3a565ff210 ("ice: add admin commands to access cgu configuration")Fixes: d7999f5ea64b ("ice: implement dpll interface to control cgu")Reported-by: kernel test robot Closes: https://lore.kernel.org/oe-kbuild-all/202309191214.TaYEct4H-lkp@intel.comSigned-off-by: Jacob Keller Reviewed-by: Przemek Kitszel Signed-off-by: Tony Nguyen Reviewed-by: Simon Horman Tested-by: Simon Horman # build-testedLink: https://lore.kernel.org/r/20231002185132.1575271-1-anthony.l.nguyen@intel.comSigned-off-by: Jakub Kicinski List of files: /linux-6.15/drivers/net/ethernet/intel/ice/Makefile

d7999f5e - ice: implement dpll interface to control cgu

Wed, 13 Sep 2023 20:49:41 +0000

ice: implement dpll interface to control cguControl over clock generation unit is required for further developmentof Synchronous Ethernet feature. Interface provides ability to obtaincurrent state of a dpll, its sources and outputs which are pins, andallows their configuration.Co-developed-by: Milena Olech Signed-off-by: Milena Olech Co-developed-by: Michal Michalik Signed-off-by: Michal Michalik Signed-off-by: Arkadiusz Kubalewski Signed-off-by: Vadim Fedorenko Signed-off-by: Jiri Pirko Signed-off-by: David S. Miller List of files: /linux-6.15/drivers/net/ethernet/intel/ice/Makefile

f6e8fb55 - ice: Implement basic eswitch bridge setup

Wed, 12 Jul 2023 11:03:31 +0000

ice: Implement basic eswitch bridge setupWith this patch, ice driver is able to track if the portrepresentors or uplink port were added to the linux bridge inswitchdev mode. Listen for NETDEV_CHANGEUPPER events in order todetect this. ice_esw_br data structure reflects the linux bridgeand stores all the ports of the bridge (ice_esw_br_port) inxarray, it's created when the first port is added to the bridge andfreed once the last port is removed. Note that only one bridge issupported per eswitch.Bridge port (ice_esw_br_port) can be either a VF port representorport or uplink port (ice_esw_br_port_type). In both cases bridge portholds a reference to the VSI, VF's VSI in case of the PR and uplinkVSI in case of the uplink. VSI's index is used as an index to thexarray in which ports are stored.Add a check which prevents configuring switchdev mode if uplink isalready added to any bridge. This is needed because we need to listenfor NETDEV_CHANGEUPPER events to record if the uplink was added tothe bridge. Netdevice notifier is registered after eswitch modeis changed to switchdev.Reviewed-by: Simon Horman Signed-off-by: Wojciech Drewek Tested-by: Sujai Buvaneswaran Signed-off-by: Tony Nguyen List of files: /linux-6.15/drivers/net/ethernet/intel/ice/Makefile

38e97a98 - ice: move interrupt related code to separate file

Mon, 15 May 2023 19:03:12 +0000

ice: move interrupt related code to separate fileKeep interrupt handling code in a dedicated file. This helps keep driverstructured better and prepares for more functionality added to this file.Reviewed-by: Jacob Keller Reviewed-by: Simon Horman Tested-by: Pucha Himasekhar Reddy (A Contingent worker at Intel)Signed-off-by: Piotr Raczynski Signed-off-by: Tony Nguyen List of files: /linux-6.15/drivers/net/ethernet/intel/ice/Makefile

cf871006 - ice: remove unnecessary CONFIG_ICE_GNSS

Fri, 24 Feb 2023 21:32:41 +0000

ice: remove unnecessary CONFIG_ICE_GNSSCONFIG_ICE_GNSS was added by commit c7ef8221ca7d ("ice: use GNSS subsysteminstead of TTY") as a way to allow the ice driver to optionally supportGNSS features without forcing a dependency on CONFIG_GNSS.The original implementation of that commit at [1] used IS_REACHABLE. Thiswas rejected by Olek at [2] with the suggested implementation ofCONFIG_ICE_GNSS.Eventually after merging, Linus reported a .config which hadCONFIG_ICE_GNSS = y when both GNSS = n and ICE = n. This confused him andhe felt that the config option was not useful, and commented about it at[3].CONFIG_ICE_GNSS is defined to y whenever GNSS = ICE. This results in itbeing set in cases where both options are not enabled.The goal of CONFIG_ICE_GNSS is to ensure that the GNSS support in the icedriver is enabled when GNSS is enabled.The complaint from Olek about the original IS_REACHABLE was due to therequired IS_REACHABLE checks throughout the ice driver code and the factthat ice_gnss.c was compiled regardless of GNSS support.This can be fixed in the Makefile by using ice-$(CONFIG_GNSS) += ice_gnss.oIn this case, if GNSS = m and ICE = y, we can result in some confusingbehavior where GNSS support is not enabled because its not built in. See[4].To disallow this, have CONFIG_ICE depend on GNSS || GNSS = n. This ensuresthat we cannot enable CONFIG_ICE as builtin while GNSS is a module.Drop CONFIG_ICE_GNSS, and replace the IS_ENABLED checks for it withchecks for GNSS. Update the Makefile to add the ice_gnss.o object based onCONFIG_GNSS.This works to ensure that GNSS support can optionally be enabled, doesn'thave an unnnecessary extra config option, and has Kbuild enforce thedependency such that you can't accidentally enable GNSS as a module and ICEas a builtin.[1] https://lore.kernel.org/intel-wired-lan/20221019095603.44825-1-arkadiusz.kubalewski@intel.com/[2] https://lore.kernel.org/intel-wired-lan/20221028165706.96849-1-alexandr.lobakin@intel.com/[3] https://lore.kernel.org/all/CAHk-=wi_410KZqHwF-WL5U7QYxnpHHHNP-3xL=g_y89XnKc-uw@mail.gmail.com/[4] https://lore.kernel.org/netdev/20230223161309.0e439c5f@kernel.org/Reported-by: Linus Torvalds Signed-off-by: Jacob Keller Fixes: c7ef8221ca7d ("ice: use GNSS subsystem instead of TTY")Cc: Arkadiusz Kubalewski Cc: Alexander Lobakin Cc: Jakub Kicinski Cc: Anthony Nguyen Acked-by: Jakub Kicinski Signed-off-by: Tony Nguyen Signed-off-by: David S. Miller List of files: /linux-6.15/drivers/net/ethernet/intel/ice/Makefile

c7ef8221 - ice: use GNSS subsystem instead of TTY

Thu, 19 Jan 2023 00:58:36 +0000

ice: use GNSS subsystem instead of TTYPreviously support for GNSS was implemented as a TTY driver, it allowedto access GNSS receiver on /dev/ttyGNSS_.Use generic GNSS subsystem API instead of implementing own TTY driver.The receiver is accessible on /dev/gnss. In case of multiple receiversin the OS, correct device can be found by enumerating either:- /sys/class/net//device/gnss/- /sys/class/gnss/gnss/device/Using GNSS subsystem is superior to implementing own TTY driver, as theGNSS subsystem was designed solely for this purpose. It also implementsTTY driver but in a common and defined way.From user perspective, there is no difference in communicating with adevice, except new path to the device shall be used. The device willprovide same information to the userspace as the old one, and can be usedin the same way, i.e.:old # gpsmon /dev/ttyGNSS_2100_0new # gpsmon /dev/gnss0There is no other impact on userspace tools.User expecting onboard GNSS receiver support is required to enableCONFIG_GNSS=y/m in kernel config.Reviewed-by: Alexander Lobakin Signed-off-by: Karol Kolacinski Signed-off-by: Michal Michalik Signed-off-by: Arkadiusz Kubalewski Tested-by: Gurucharan G (A Contingent worker at Intel)Signed-off-by: Tony Nguyen Acked-by: Greg Kroah-Hartman Signed-off-by: David S. Miller List of files: /linux-6.15/drivers/net/ethernet/intel/ice/Makefile

2ffd87d3 - ice: Move support DDP code out of ice_flex_pipe.c

Tue, 20 Dec 2022 15:01:14 +0000

ice: Move support DDP code out of ice_flex_pipe.cCurrently, ice_flex_pipe.c includes the DDP loading functionsand has grown large. Although flexible processing supportcode is related to DDP loading, these parts are distinct.Move the DDP loading functionality from ice_flex_pipe.c toa separate file.Signed-off-by: Sergey Temerkhanov Tested-by: Gurucharan G (A Contingent worker at Intel)Signed-off-by: Tony Nguyen List of files: /linux-6.15/drivers/net/ethernet/intel/ice/Makefile

6e1ff618 - ice: fix access-beyond-end in the switch code

Wed, 01 Jun 2022 10:59:24 +0000

ice: fix access-beyond-end in the switch codeGlobal `-Warray-bounds` enablement revealed some problems, one ofwhich is the way we define and use AQC rules messages.In fact, they have a shared header, followed by the actual message,which can be of one of several different formats. So it isstraightforward enough to define that header as a separate structand then embed it into message structures as needed, but currentlyall the formats reside in one union coupled with the header. Then,the code allocates only the memory needed for a particular messageformat, leaving the union potentially incomplete.There are no actual reads or writes beyond the end of an allocatedchunk, but at the same time, the whole implementation is fragile andbacked by an equilibrium rather than strong type and memory checks.Define the structures the other way around: one for the commonheader and the rest for the actual formats with the header embedded.There are no places where several union members would be used at thesame time anyway. This allows to use proper struct_size() and letthe compiler know what is going to be done.Finally, unsilence `-Warray-bounds` back for ice_switch.c.Other little things worth mentioning:* &ice_sw_rule_vsi_list_query is not used anywhere, remove it. It's weird anyway to talk to hardware with purely kernel types (bitmaps);* expand the ICE_SW_RULE_*_SIZE() macros to pass a structure variable name to struct_size() to let it do strict typechecking;* rename ice_sw_rule_lkup_rx_tx::hdr to ::hdr_data to keep ::hdr for the header structure to have the same name for it constistenly everywhere;* drop the duplicate of %ICE_SW_RULE_RX_TX_NO_HDR_SIZE residing in ice_switch.h.Fixes: 9daf8208dd4d ("ice: Add support for switch filter programming")Fixes: 66486d8943ba ("ice: replace single-element array used for C struct hack")Signed-off-by: Alexander Lobakin Reviewed-by: Marcin Szycik Acked-by: Tony Nguyen Link: https://lore.kernel.org/r/20220601105924.2841410-1-alexandr.lobakin@intel.comSigned-off-by: Paolo Abeni List of files: /linux-6.15/drivers/net/ethernet/intel/ice/Makefile

385bc51b - eth: ice: silence the GCC 12 array-bounds warning

Fri, 20 May 2022 19:56:04 +0000

eth: ice: silence the GCC 12 array-bounds warningGCC 12 gets upset because driver allocates partialstruct ice_aqc_sw_rules_elem buffers. The writes arewithin bounds.Silence these warnings for now, our build bot runs GCC 12so we won't allow any new instances.Signed-off-by: Jakub Kicinski Signed-off-by: David S. Miller List of files: /linux-6.15/drivers/net/ethernet/intel/ice/Makefile