ETHTOOL(8)

📦
version 7.2 · 11 versions
, in centos-stream 10, fedora 44

👑 ETHTOOL(8) — System Manager's Manual 👑

okay so bestie, this is the manual for ethtool, and it is literally the main character of wired Ethernet tweaking. no cap. 💅

🌸 NAME 🌸

ethtool — query or control network driver and hardware settings.

✨ SYNOPSIS ✨

here's the whole shape of it, bestie — every invocation form, exactly as it lives:

ethtool devname

ethtool -h|--help

ethtool --version

ethtool [--debug N] args

ethtool [--disable-netlink] args

ethtool [--json] args

ethtool [-I | --include-statistics] args

ethtool [--phy N] args

ethtool --monitor [ command ] [ devname ]

ethtool -a|--show-pause devname

ethtool -A|--pause devname [autoneg on|off] [rx on|off] [tx on|off]

ethtool -c|--show-coalesce devname

ethtool -C|--coalesce devname [adaptive-rx on|off] [adaptive-tx on|off]
       [rx-usecs N] [rx-frames N] [rx-usecs-irq N] [rx-frames-irq N]
       [tx-usecs N] [tx-frames N] [tx-usecs-irq N] [tx-frames-irq N]
       [stats-block-usecs N] [pkt-rate-low N] [rx-usecs-low N]
       [rx-frames-low N] [tx-usecs-low N] [tx-frames-low N]
       [pkt-rate-high N] [rx-usecs-high N] [rx-frames-high N]
       [tx-usecs-high N] [tx-frames-high N] [sample-interval N]
       [cqe-mode-rx on|off] [cqe-mode-tx on|off] [tx-aggr-max-bytes N]
       [tx-aggr-max-frames N] [tx-aggr-time-usecs N] [rx-cqe-frames N]
       [rx-cqe-nsecs N]

ethtool -g|--show-ring devname

ethtool -G|--set-ring devname [rx N] [rx-mini N] [rx-jumbo N] [tx N]
       [rx-buf-len N] [tcp-data-split auto|on|off] [hds-thresh N]
       [cqe-size N] [tx-push N] [rx-push N] [tx-push-buf-len N]

ethtool -i|--driver devname

ethtool -d|--register-dump devname [raw on|off] [hex on|off] [file name]

ethtool -e|--eeprom-dump devname [raw on|off] [offset N] [length N]

ethtool -E|--change-eeprom devname [magic N] [offset N] [length N]
       [value N]

ethtool -k|--show-features|--show-offload devname

ethtool -K|--features|--offload devname feature on|off ...

ethtool -p|--identify devname [N]

ethtool -P|--show-permaddr devname

ethtool -r|--negotiate devname

ethtool -S|--statistics devname [--all-groups|--groups [eth-phy]
       [eth-mac] [eth-ctrl] [rmon] ]

ethtool --phy-statistics devname

ethtool -t|--test devname [offline|online|external_lb]

ethtool -s devname [speed N] [lanes N] [duplex half|full]
       [port tp|aui|bnc|mii] [mdix auto|on|off] [autoneg on|off]
       [advertise N[/M] | advertise mode on|off ...]  [phyad N]
       [xcvr internal|external] [wol N[/M] | wol p|u|m|b|a|g|s|f|d...]
       [sopass xx:yy:zz:aa:bb:cc] [master-slave preferred-master|preferred-slave|forced-master|forced-slave] [msglvl
       N[/M] | msglvl type on|off ...]

ethtool -n|-u|--show-nfc|--show-ntuple devname
       [ rx-flow-hash tcp4|udp4|ah4|esp4|sctp4|tcp6|udp6|ah6|esp6|sctp6 |
       rule N ]

ethtool -N|-U|--config-nfc|--config-ntuple devname
       rx-flow-hash tcp4|udp4|ah4|esp4|sctp4|tcp6|udp6|ah6|esp6|sctp6
       m|v|t|s|d|f|n|r... |
       flow-type
       ether|ip4|tcp4|udp4|sctp4|ah4|esp4|ip6|tcp6|udp6|ah6|esp6|sctp6
       [src xx:yy:zz:aa:bb:cc [m xx:yy:zz:aa:bb:cc]]
       [dst xx:yy:zz:aa:bb:cc [m xx:yy:zz:aa:bb:cc]] [proto N [m N]]
       [src-ip ip-address [m ip-address]] [dst-ip ip-address [m ip-address]] [tos N [m N]] [tclass N [m N]] [l4proto N [m N]]
       [src-port N [m N]] [dst-port N [m N]] [spi N [m N]]
       [l4data N [m N]] [vlan-etype N [m N]] [vlan N [m N]]
       [user-def N [m N]] [dst-mac xx:yy:zz:aa:bb:cc [m xx:yy:zz:aa:bb:cc]] [action N]
       [context N] [loc N] |
       delete N

ethtool -w|--get-dump devname [data filename]

ethtool -W|--set-dump devname N

ethtool -T|--show-time-stamping devname [index N qualifier precise|approx]

ethtool --get-hwtimestamp-cfg devname

ethtool --set-hwtimestamp-cfg devname [index N qualifier precise|approx]
       [tx TX-TYPE] [rx-filter RX-FILTER]

ethtool -x|--show-rxfh-indir|--show-rxfh devname

ethtool -X|--set-rxfh-indir|--rxfh devname [hkey xx:yy:zz:aa:bb:cc:...]
       [start N] [ equal N | weight W0 W1 ... | default ] [hfunc FUNC]
       [xfrm symmetric-xor|symmetric-or-xor|none] [context CTX | new]
       [delete]

ethtool -f|--flash devname file [N]

ethtool -l|--show-channels devname

ethtool -L|--set-channels devname [rx N] [tx N] [other N] [combined N]

ethtool -m|--dump-module-eeprom|--module-info devname [raw on|off]
       [hex on|off] [pages on|off] [offset N] [length N] [page N]
       [bank N] [i2c N]

ethtool --show-priv-flags devname

ethtool --set-priv-flags devname flag on|off ...

ethtool --show-eee devname

ethtool --set-eee devname [eee on|off] [tx-lpi on|off] [tx-timer N]
       [advertise N]

ethtool --set-phy-tunable devname [ downshift on|off [count N] ] [
       fast-link-down on|off [msecs N] ] [ energy-detect-power-down
       on|off [msecs N] ] [ short-cable-preset on|off ] [ lpf-bandwidth
       [value N] ] [ dsp-eq-init-value [value N] ]

ethtool --get-phy-tunable devname [downshift] [fast-link-down] [energy-detect-power-down] [short-cable-preset] [lpf-bandwidth] [dsp-eq-init-value]

ethtool --get-tunable devname [rx-copybreak] [tx-copybreak] [tx-buf-size] [pfc-prevention-tout]

ethtool --set-tunable devname [rx-copybreak N] [tx-copybreak N]
       [tx-buf-size N] [pfc-prevention-tout N]

ethtool --reset devname [flags N] [mgmt] [mgmt-shared] [irq] [irq-shared] [dma] [dma-shared] [filter] [filter-shared] [offload]
       [offload-shared] [mac] [mac-shared] [phy] [phy-shared] [ram]
       [ram-shared] [ap] [ap-shared] [dedicated] [all]

ethtool --show-fec devname

ethtool --set-fec devname encoding auto|off|rs|baser|llrs [...]

ethtool -Q|--per-queue devname [queue_mask %x] sub_command ...

ethtool --cable-test devname

ethtool --cable-test-tdr devname [first N] [last N] [step N] [pair N]

ethtool --show-tunnels devname

ethtool --show-module devname

ethtool --set-module devname [power-mode-policy high|auto]

ethtool --get-plca-cfg devname

ethtool --set-plca-cfg devname [enable on|off] [node-id N] [node-cnt N]
       [to-tmr N] [burst-cnt N] [burst-tmr N]

ethtool --get-plca-status devname

ethtool --show-mm devname

ethtool --set-mm devname [verify-enabled on|off] [verify-time N]
       [tx-enabled on|off] [pmac-enabled on|off] [tx-min-frag-size N]

ethtool --show-pse devname

ethtool --set-pse devname [podl-pse-admin-control enable|disable]
       [c33-pse-admin-control enable|disable] [c33-pse-avail-pw-limit N]
       [prio N]

ethtool --flash-module-firmware devname file FILE [pass PASS]

ethtool --show-phys devname

🦋 DESCRIPTION 🦋

okay so ethtool is the tool you use to query and control network device driver and hardware settings, especially for wired Ethernet devices. it's giving hardware whisperer, honestly. 🔥

and devname? that's literally the name of the network device on which ethtool should operate. periodt. 💅

🎀 OPTIONS 🎀

when you call ethtool with a single argument that specifies the device name, it prints the current settings of that device. cute, useful, iconic.

  • -h, --help — shows a short help message. your lil rescue text, bestie. 🫶
  • --version — shows the ethtool version number.
  • --debug N — turns on debugging messages. the argument is interpreted as a mask:
    • 0x01 — parser information
    • 0x02 — summary of netlink messages
    • 0x04 — hex dump of sent netlink messages
    • 0x08 — hex dump of received netlink messages
    • 0x10 — structure of netlink messages
  • --disable-netlink — do not use netlink, and fall back to the ioctl interface if possible. lowkey old-school mode. ✨
  • --json — output results in JavaScript Object Notation (JSON). only a subset of options support this, bestie. those which do not will continue to output plain text in the presence of this option.
  • -I, --include-statistics — include command-related statistics in the output. this option allows displaying relevant device statistics for selected get commands.
  • --phy N — target a PHY within the interface. the PHY index can be retrieved with --show-phys. PHY index 0 targets the phy device directly attached to the ethernet MAC, if any. the following commands can accept a PHY index:
    • --cable-test
    • --cable-test-tdr
    • --get-plca-cfg
    • --set-plca-cfg
    • --get-plca-status
    • --show-pse
    • --set-pse
    • --show-mse
  • -a, --show-pause — queries the specified Ethernet device for pause parameter information.
    • --src aggregate|emac|pmac — if the MAC Merge layer is supported, request a particular source of device statistics (eMAC or pMAC, or their aggregate). only valid if ethtool was invoked with the -I --include-statistics argument.
  • -A, --pause — changes the pause parameters of the specified Ethernet device.
    • autoneg on|off — specifies whether pause autonegotiation should be enabled.
    • rx on|off — specifies whether RX pause should be enabled.
    • tx on|off — specifies whether TX pause should be enabled.
  • -c, --show-coalesce — queries the specified network device for coalescing information.
  • -C, --coalesce — changes the coalescing settings of the specified network device.
  • -g, --show-ring — queries the specified network device for rx/tx ring parameter information.
  • -G, --set-ring — changes the rx/tx ring parameters of the specified network device.
    • rx N — changes the number of ring entries for the Rx ring.
    • rx-mini N — changes the number of ring entries for the Rx Mini ring.
    • rx-jumbo N — changes the number of ring entries for the Rx Jumbo ring.
    • tx N — changes the number of ring entries for the Tx ring.
    • rx-buf-len N — changes the size of a buffer in the Rx ring.
    • tcp-data-split auto|on|off — specifies the state of TCP data split.
    • hds-thresh N — specifies the threshold value of tcp data split.
    • cqe-size N — changes the size of completion queue event.
    • tx-push on|off — specifies whether TX push should be enabled.
    • rx-push on|off — specifies whether RX push should be enabled.
    • tx-push-buf-len N — specifies the maximum number of bytes of a transmitted packet a driver can push directly to the underlying device.
  • -i, --driver — queries the specified network device for associated driver information.
  • -d, --register-dump — bestie, this one grabs and prints a register dump for the network device you picked. Some devices have a known register format and get decoded; others just print in hex. When raw is enabled, then ethtool dumps the raw register data to stdout. If file is specified, then it uses the contents of a previous raw register dump instead of reading from the device. 💅
  • -e, --eeprom-dump — retrieves and prints an EEPROM dump for the specified network device. When raw is enabled, then it dumps the raw EEPROM data to stdout. The length and offset parameters let you dump certain portions of the EEPROM. Default is to dump the entire EEPROM. 🎀
    • raw on|off — turns raw mode on or off for the dump.
    • offset N — the offset knob.
    • length N — the length knob.
  • -E, --change-eeprom — if value is specified, changes the EEPROM byte for the specified network device. offset and value specify which byte and its new value. If value is not specified, stdin is read and written to the EEPROM. The length and offset parameters allow writing to certain portions of the EEPROM.
⚠️ bestie, because of the persistent nature of writing to the EEPROM, a device-specific magic key must be specified to prevent the accidental writing to the EEPROM. like, respect the magic key fr fr.
  • -k, --show-features, --show-offload — queries the specified network device for the state of protocol offload and other features. 👑
  • -K, --features, --offload — changes the offload parameters and other features of the specified network device. The following feature names are built-in and others may be defined by the kernel. ✨
    • rx on|off — specifies whether RX checksumming should be enabled.
    • tx on|off — specifies whether TX checksumming should be enabled.
    • sg on|off — specifies whether scatter-gather should be enabled.
    • tso on|off — specifies whether TCP segmentation offload should be enabled.
    • ufo on|off — specifies whether UDP fragmentation offload should be enabled.
    • gso on|off — specifies whether generic segmentation offload should be enabled.
    • gro on|off — specifies whether generic receive offload should be enabled.
    • lro on|off — specifies whether large receive offload should be enabled.
    • rxvlan on|off — specifies whether RX VLAN acceleration should be enabled.
    • txvlan on|off — specifies whether TX VLAN acceleration should be enabled.
    • ntuple on|off — specifies whether Rx ntuple filters and actions should be enabled.
    • rxhash on|off — specifies whether receive hashing offload should be enabled.
  • -p, --identify — initiates adapter-specific action intended to enable an operator to easily identify the adapter by sight. Typically this involves blinking one or more LEDs on the specific network port. 🦋
    • [ N] — length of time to perform phys-id, in seconds.
  • -P, --show-permaddr — queries the specified network device for permanent hardware address. 🫶
  • -r, --negotiate — restarts auto-negotiation on the specified Ethernet device, if auto-negotiation is enabled. 🌸
  • -S, --statistics — queries the specified network device for standard (IEEE, IETF, etc.), or NIC- and driver-specific statistics. NIC- and driver-specific statistics are requested when no group of statistics is specified. 😭 NIC- and driver-specific statistics and standard statistics are independent, devices may implement either, both or none. There is little commonality between naming of NIC- and driver-specific statistics across vendors. lowkey chaotic, but that's the vibe.
    • --all-groups — no extra description given here, bestie, it's just the named flag living its truth.
    • --groups [eth-phy] [eth-mac] [eth-ctrl] [rmon] — request groups of standard device statistics.
    • --src aggregate|emac|pmac — if the MAC Merge layer is supported, request a particular source of device statistics (eMAC or pMAC, or their aggregate).
  • --phy-statistics — queries the specified network device for PHY specific statistics. 💜
  • -t, --test — executes adapter selftest on the specified network device. Possible test modes are: 🧚
    • offline — perform full set of tests, possibly interrupting normal operation during the tests.
    • online — perform limited set of tests, not interrupting normal operation.
    • external_lb — perform full set of tests, as for offline, and additionally an external-loopback test.
  • -s, --change — allows changing some or all settings of the specified network device. All following options only apply if -s was specified. main character energy, honestly. 🌈
    • speed N — set speed in Mb/s. ethtool with just the device name as an argument will show you the supported device speeds.
    • lanes N — set number of lanes.
    • duplex half|full — sets full or half duplex mode.
    • port tp|aui|bnc|mii — selects device port.
    • master-slave preferred-master|preferred-slave|forced-master|forced-slave — configure MASTER/SLAVE role of the PHY. When the PHY is configured as MASTER, the PMA Transmit function shall source TX_TCLK from a local clock source. When configured as SLAVE, the PMA Transmit function shall source TX_TCLK from the clock recovered from data stream provided by MASTER. Not all devices support this. 🫶
      • preferred-master — prefer MASTER role on autonegotiation.
      • preferred-slave — prefer SLAVE role on autonegotiation.
      • forced-master — force the PHY in MASTER role. Can be used without autonegotiation.
      • forced-slave — force the PHY in SLAVE role. Can be used without autonegotiation.
    • mdix auto|on|off — selects MDI-X mode for port. May be used to override the automatic detection feature of most adapters. An argument of auto means automatic detection of MDI status, on forces MDI-X (crossover) mode, while off means MDI (straight through) mode. The driver should guarantee that this command takes effect immediately, and if necessary may reset the link to cause the change to take effect. 💅
    • autoneg on|off — specifies whether autonegotiation should be enabled. Autonegotiation is enabled by default, but in some network devices may have trouble with it, so you can disable it if really necessary. 🍓
  • advertise N — tells autonegotiation which speeds and duplex modes to brag about 💅 N is a hexadecimal value, and you can OR together as many of the following as you want. Here's the whole menu, bestie — yes, all of it, I would never gatekeep:
0x001                                10baseT Half
0x002                                10baseT Full
0x100000000000000000000000           10baseT1L Full
0x8000000000000000000000000          10baseT1S Full
0x10000000000000000000000000         10baseT1S Half
0x20000000000000000000000000         10baseT1S_P2MP Half
0x40000000000000000000000000         10baseT1BRR Full
0x004                                100baseT Half
0x008                                100baseT Full
0x80000000000000000                  100baseT1 Full
0x40000000000000000000000            100baseFX Half
0x80000000000000000000000            100baseFX Full
0x010                                1000baseT Half               (not supported by IEEE standards)
0x020                                1000baseT Full
0x20000                              1000baseKX Full
0x20000000000                        1000baseX Full
0x100000000000000000                 1000baseT1 Full
0x8000                               2500baseX Full               (not supported by IEEE standards)
0x800000000000                       2500baseT Full
0x1000000000000                      5000baseT Full
0x1000                               10000baseT Full
0x40000                              10000baseKX4 Full
0x80000                              10000baseKR Full
0x100000                             10000baseR_FEC
0x40000000000                        10000baseCR Full             (a.k.a 10G_SFI_DA, SFF-8431 Appendix-E)
0x80000000000                        10000baseSR Full
0x100000000000                       10000baseLR Full
0x200000000000                       10000baseLRM Full
0x400000000000                       10000baseER Full
0x200000                             20000baseMLD2 Full           (not supported by IEEE standards)
0x400000                             20000baseKR2 Full            (not supported by IEEE standards)
0x80000000                           25000baseCR Full
0x100000000                          25000baseKR Full
0x200000000                          25000baseSR Full
0x800000                             40000baseKR4 Full
0x1000000                            40000baseCR4 Full
0x2000000                            40000baseSR4 Full
0x4000000                            40000baseLR4 Full
0x400000000                          50000baseCR2 Full
0x800000000                          50000baseKR2 Full
0x10000000000                        50000baseSR2 Full
0x10000000000000                     50000baseKR Full
0x20000000000000                     50000baseSR Full
0x40000000000000                     50000baseCR Full
0x80000000000000                     50000baseLR_ER_FR Full
0x100000000000000                    50000baseDR Full
0x8000000                            56000baseKR4 Full
0x10000000                           56000baseCR4 Full
0x20000000                           56000baseSR4 Full
0x40000000                           56000baseLR4 Full
0x1000000000                         100000baseKR4 Full
0x2000000000                         100000baseSR4 Full
0x4000000000                         100000baseCR4 Full
0x8000000000                         100000baseLR4_ER4 Full
0x200000000000000                    100000baseKR2 Full
0x400000000000000                    100000baseSR2 Full
0x800000000000000                    100000baseCR2 Full
0x1000000000000000                   100000baseLR2_ER2_FR2 Full
0x2000000000000000                   100000baseDR2 Full
0x8000000000000000000                100000baseKR Full
0x10000000000000000000               100000baseSR Full
0x20000000000000000000               100000baseLR_ER_FR Full
0x40000000000000000000               100000baseCR Full
0x80000000000000000000               100000baseDR Full
0x4000000000000000                   200000baseKR4 Full
0x8000000000000000                   200000baseSR4 Full
0x10000000000000000                  200000baseLR4_ER4_FR4 Full
0x20000000000000000                  200000baseDR4 Full
0x40000000000000000                  200000baseCR4 Full
0x100000000000000000000              200000baseKR2 Full
0x200000000000000000000              200000baseSR2 Full
0x400000000000000000000              200000baseLR2_ER2_FR2 Full
0x800000000000000000000              200000baseDR2 Full
0x1000000000000000000000             200000baseCR2 Full
0x80000000000000000000000000         200000baseCR Full
0x100000000000000000000000000        200000baseKR Full
0x200000000000000000000000000        200000baseDR Full
0x400000000000000000000000000        200000baseDR_2 Full
0x800000000000000000000000000        200000baseSR Full
0x1000000000000000000000000000       200000baseVR Full
0x200000000000000000                 400000baseKR8 Full
0x400000000000000000                 400000baseSR8 Full
0x800000000000000000                 400000baseLR8_ER8_FR8 Full
0x1000000000000000000                400000baseDR8 Full
0x2000000000000000000                400000baseCR8 Full
0x2000000000000000000000             400000baseKR4 Full
0x4000000000000000000000             400000baseSR4 Full
0x8000000000000000000000             400000baseLR4_ER4_FR4 Full
0x10000000000000000000000            400000baseDR4 Full
0x20000000000000000000000            400000baseCR4 Full
0x2000000000000000000000000000       400000baseCR2 Full
0x4000000000000000000000000000       400000baseKR2 Full
0x8000000000000000000000000000       400000baseDR2 Full
0x10000000000000000000000000000      400000baseDR2_2 Full
0x20000000000000000000000000000      400000baseSR2 Full
0x40000000000000000000000000000      400000baseVR2 Full
0x200000000000000000000000           800000baseCR8 Full
0x400000000000000000000000           800000baseKR8 Full
0x800000000000000000000000           800000baseDR8 Full
0x1000000000000000000000000          800000baseDR8_2 Full
0x2000000000000000000000000          800000baseSR8 Full
0x4000000000000000000000000          800000baseVR8 Full
0x80000000000000000000000000000      800000baseCR4 Full
0x100000000000000000000000000000     800000baseKR4 Full
0x200000000000000000000000000000     800000baseDR4 Full
0x400000000000000000000000000000     800000baseDR4_2 Full
0x800000000000000000000000000000     800000baseSR4 Full
0x1000000000000000000000000000000    800000baseVR4 Full
0x2000000000000000000000000000000    1600000baseCR8 Full
0x4000000000000000000000000000000    1600000baseKR8 Full
0x8000000000000000000000000000000    1600000baseDR8 Full
0x10000000000000000000000000000000   1600000baseDR8_2 Full
bestie please notice the little "(not supported by IEEE standards)" notes on a few of these — they're still listed, they're just spicy non-standard vibes, so don't come for me 🌸
  • phyad N — sets the PHY address. That's it, that's the whole thing 🎀
  • xcvr internal|external — picks the transceiver type. Right now only internal and external are valid; in the future further types might be added, so keep an eye out 👀
  • wol p|u|m|b|a|g|s|f|d... — sets Wake-on-LAN options. Not all devices support this, so vibe check your hardware first 😭 the argument is a string of characters saying which options to turn on:
    • p — Wake on PHY activity
    • u — Wake on unicast messages
    • m — Wake on multicast messages
    • b — Wake on broadcast messages
    • a — Wake on ARP
    • g — Wake on MagicPacket™
    • s — Enable SecureOn™ password for MagicPacket™
    • f — Wake on filter(s)
    • d — Disable (wake on nothing). This option clears all previous options.
  • sopass xx:yy:zz:aa:bb:cc — sets the SecureOn™ password. The argument has to be exactly 6 bytes in Ethernet MAC hex format (xx:yy:zz:aa:bb:cc) — no improvising, she's strict 💅
  • msglvl N
  • msglvl type on|off ... — sets the driver message type flags, either by name or by number. type names the type of message to enable or disable; N specifies the new flags numerically. The defined type names and numbers are:
    • drv — 0x0001 — General driver status
    • probe — 0x0002 — Hardware probing
    • link — 0x0004 — Link state
    • timer — 0x0008 — Periodic status check
    • ifdown — 0x0010 — Interface being brought down
    • ifup — 0x0020 — Interface being brought up
    • rx_err — 0x0040 — Receive error
    • tx_err — 0x0080 — Transmit error
    • tx_queued — 0x0100 — Transmit queueing
    • intr — 0x0200 — Interrupt handling
    • tx_done — 0x0400 — Transmit completion
    • rx_status — 0x0800 — Receive completion
    • pktdata — 0x1000 — Packet contents
    • hw — 0x2000 — Hardware status
    • wol — 0x4000 — Wake-on-LAN status

The precise meanings of these type flags differ between drivers — so like, don't assume, iykyk 🦋

  • -n, -u, --show-nfc, --show-ntuple — retrieves receive network flow classification options or rules. getting the receipts on your flow rules ✨
  • rx-flow-hash tcp4|udp4|ah4|esp4|sctp4|tcp6|udp6|ah6|esp6|sctp6 — pulls the hash options for the flow type you name:
    • tcp4 — TCP over IPv4
    • udp4 — UDP over IPv4
    • ah4 — IPSEC AH over IPv4
    • esp4 — IPSEC ESP over IPv4
    • sctp4 — SCTP over IPv4
    • tcp6 — TCP over IPv6
    • udp6 — UDP over IPv6
    • ah6 — IPSEC AH over IPv6
    • esp6 — IPSEC ESP over IPv6
    • sctp6 — SCTP over IPv6
  • rule N — retrieves the RX classification rule with the given ID. just hand her the ID and she'll hand you the rule, no cap 👑
  • -N, -U, --config-nfc, --config-ntuple — sets up your receive network flow classification options or rules, bestie. This is the "tell the NIC exactly which packets go where" button 💅

the rx-flow-hash situation 🌀

rx-flow-hash tcp4|udp4|ah4|esp4|sctp4|tcp6|udp6|ah6|esp6|sctp6
m|v|t|s|d|f|n|r...

Configures the hash options for whichever flow type you named. The letters mean:

  • m — hash on the Layer 2 destination address of the rx packet.
  • v — hash on the VLAN tag of the rx packet.
  • t — hash on the Layer 3 protocol field of the rx packet.
  • s — hash on the IP source address of the rx packet.
  • d — hash on the IP destination address of the rx packet.
  • f — hash on bytes 0 and 1 of the Layer 4 header of the rx packet.
  • n — hash on bytes 2 and 3 of the Layer 4 header of the rx packet.
  • l — hash on IPv6 Flow Label of the rx packet.
  • r — discard all packets of this flow type. When this one's set, all the other options are ignored, periodt.

the flow-type moment 🦋

ether|ip4|tcp4|udp4|sctp4|ah4|esp4|ip6|tcp6|udp6|ah6|esp6|sctp6

Inserts or updates a classification rule for the flow type you picked:

  • ether — Ethernet
  • ip4 — Raw IPv4
  • tcp4 — TCP over IPv4
  • udp4 — UDP over IPv4
  • sctp4 — SCTP over IPv4
  • ah4 — IPSEC AH over IPv4
  • esp4 — IPSEC ESP over IPv4
  • ip6 — Raw IPv6
  • tcp6 — TCP over IPv6
  • udp6 — UDP over IPv6
  • sctp6 — SCTP over IPv6
  • ah6 — IPSEC AH over IPv6
  • esp6 — IPSEC ESP over IPv6
Okay so here's the mask tea ☕ — for all fields that let you specify both a value and a mask, you can drop the mask immediately after the value using the m keyword, or give it separately using the field name keyword with -mask appended, like src-mask. Two ways, same energy.

the actual fields 🎀

  • src xx:yy:zz:aa:bb:cc [m xx:yy:zz:aa:bb:cc] — includes the source MAC address (6 bytes, hex, colon-separated) plus an optional mask. Only valid for flow-type ether.
  • dst xx:yy:zz:aa:bb:cc [m xx:yy:zz:aa:bb:cc] — includes the destination MAC address (6 bytes, hex, colon-separated) plus an optional mask. Only valid for flow-type ether.
  • proto N [m N] — includes the Ethernet protocol number (ethertype) and an optional mask. Only valid for flow-type ether.
  • src-ip ip-address [m ip-address] — the source IP address of the incoming packet to match, plus an optional mask. Valid for all IP based flow-types.
  • dst-ip ip-address [m ip-address] — the destination IP address of the incoming packet to match, plus an optional mask. Valid for all IP based flow-types.
  • tos N [m N] — the value of the Type of Service field in the incoming packet to match, plus an optional mask. Applies to all IPv4 based flow-types.
  • tclass N [m N] — the value of the Traffic Class field in the incoming packet to match, plus an optional mask. Applies to all IPv6 based flow-types.
  • l4proto N [m N] — includes the layer 4 protocol number and optional mask. Valid only for flow-types ip4 and ip6.
  • src-port N [m N] — the value of the source port field (applicable to TCP/UDP packets) in the incoming packet to match, plus an optional mask. Valid for flow-types ip4, tcp4, udp4, and sctp4 and their IPv6 equivalents.
  • dst-port N [m N] — the value of the destination port field (applicable to TCP/UDP packets) in the incoming packet to match, plus an optional mask. Valid for flow-types ip4, tcp4, udp4, and sctp4 and their IPv6 equivalents.
  • spi N [m N] — the value of the security parameter index field (applicable to AH/ESP packets) in the incoming packet to match, plus an optional mask. Valid for flow-types ip4, ah4, and esp4 and their IPv6 equivalents.
  • l4data N [m N] — the value of the first 4 Bytes of Layer 4 in the incoming packet to match, plus an optional mask. Valid for ip4 and ip6 flow-types.
  • vlan-etype N [m N] — includes the VLAN tag Ethertype and an optional mask.
  • vlan N [m N] — includes the VLAN tag and an optional mask.
  • user-def N [m N] — includes 64-bits of user-specific data and an optional mask.
  • dst-mac xx:yy:zz:aa:bb:cc [m xx:yy:zz:aa:bb:cc] — includes the destination MAC address (6 bytes, hex, colon-separated) plus an optional mask. Valid for all IP based flow-types.

where does it GO tho 🚀

  • action N — specifies the Rx queue to send packets to, or some other action:
    • -1 — drop the matched flow
    • -2 — use the matched flow as a Wake-on-LAN filter
    • 0 or higher — Rx queue to route the flow
  • context N — specifies the RSS context to spread packets over multiple queues; either 0 for the default RSS context, or a value returned by ethtool -X ... context new. If you combine it with action or queue, the Rx queue you specified gets added to the value read from the RSS indirection table to produce the actual Rx queue to deliver to. iykyk 💫
  • vf N — specifies the Virtual Function the filter applies to. Not compatible with action.
  • queue N — specifies the Rx queue to send packets to. Not compatible with action.
  • loc N — specify the location/ID to insert the rule. Heads up: this will overwrite any rule present in that location and will not go through any of the rule ordering process. no cap, it just slams it in there 😭
  • delete N — deletes the RX classification rule with the given ID.
  • -w, --get-dump — retrieves and prints firmware dump for the specified network device. By default it prints out the dump flag, version and length of the dump data. When data is indicated, then ethtool fetches the dump data and directs it to a file. cute little fetch quest 🎀
  • -W, --set-dump — sets the dump flag for the device.
  • -T, --show-time-stamping — shows the device's time stamping capabilities and associated PTP hardware clock. If you don't provide an index and qualifier, it shows the capabilities of the currently selected time stamping. ⏰✨
    • index N — index of the ptp hardware clock
    • qualifier precise|approx — qualifier of the ptp hardware clock. Mainly "precise" — the default one — is for IEEE 1588 quality, and "approx" is for NICs DMA point.
  • --get-hwtimestamp-cfg — shows the selected time stamping PTP hardware clock configuration.
  • --set-hwtimestamp-cfg — sets the device's time stamping PTP hardware clock configuration.
    • index N — index of the ptp hardware clock
    • qualifier precise|approx — qualifier of the ptp hardware clock. Mainly "precise" — the default one — is for IEEE 1588 quality, and "approx" is for NICs DMA point.
    • tx TX-TYPE — type of TX time stamping to configure
    • rx-filter RX-FILTER — type of RX time stamping filter to configure
  • -x, --show-rxfh-indir, --show-rxfh — retrieves the receive flow hash indirection table and/or RSS hash key. peek at the spread 🌈
  • -X, --set-rxfh-indir, --rxfh — configures the receive flow hash indirection table and/or RSS hash key. now you're the one deciding the spread 💅
    • hkey — sets RSS hash key of the specified network device. RSS hash key should be of device supported length. Hash key format must be in xx:yy:zz:aa:bb:cc format, meaning both the nibbles of a byte should be mentioned even if a nibble is zero. don't skip the zeros bestie, it will not slay 🫶
    • hfunc — sets RSS hash function of the specified network device. The list of RSS hash functions which the kernel supports is shown as a part of the --show-rxfh command output.
    • xfrm — sets the RSS input transformation. Currently, only the symmetric-xor and symmetric-or-xor transformations are supported, where the NIC XORs/ORs the L3 and/or L4 source and destination fields (as selected by --config-nfc rx-flow-hash) before passing them to the hash algorithm. The RSS hash function will then yield the same hash for the other flow direction where the source and destination fields are swapped (i.e. Symmetric RSS).
⚠️ Real talk: this operation reduces the entropy of the input set and the hash algorithm could potentially be exploited. Switch off by xfrm none. Not me warning you twice 💀
  • start N — for the equal and weight options, sets the starting receive queue for spreading flows to N.
  • equal N — sets the receive flow hash indirection table to spread flows evenly between the first N receive queues. fairness icon ✨
  • weight W0 W1 ... — sets the receive flow hash indirection table to spread flows between receive queues according to the given weights. The sum of the weights must be non-zero and must not exceed the size of the indirection table.
  • default — sets the receive flow hash indirection table to its default value. reset era 🌸
  • context CTX | new — specifies an RSS context to act on; either new to allocate a new RSS context, or CTX, a value returned by a previous ... context new.
  • delete — delete the specified RSS context. May only be used in conjunction with context and a non-zero CTX value.
  • -f, --flash — write a firmware image to flash or other non-volatile memory on the device. glow-up time for your NIC 🔥
    • file — specifies the filename of the firmware image. The firmware must first be installed in one of the directories where the kernel firmware loader or firmware agent will look, such as /lib/firmware. so like, put it where it can be found first, okay? 🎀
    • N — if the device stores multiple firmware images in separate regions of non-volatile memory, this parameter may be used to specify which region is to be written. The default is 0, requesting that all regions are written. All other values are driver-dependent.
  • -l, --show-channels — asks the NIC bestie "so how many channels u got?" and reports back. A channel is literally an IRQ plus the whole little squad of queues that can trigger it. 💅
  • -L, --set-channels — actually changes the channel counts on the device. main character energy, we're rearranging things now:
    • rx N — changes the number of channels that have only receive queues.
    • tx N — changes the number of channels that have only transmit queues.
    • other N — changes the number of channels used for other purposes only, like link interrupts or SR-IOV co-ordination. iykyk.
    • combined N — changes the number of multi-purpose channels. the versatile ones. 🦋
  • -m, --dump-module-eeprom, --module-info — grabs the EEPROM out of plugin modules (e.g. SFP+, QSFP) and decodes it if it can. If the driver and the module both support it, the optical diagnostic info gets read and decoded too ✨. A few things to know, bestie, bc this one's got lore:
    • When either one of page, bank or i2c parameters is specified, only dumps of a single page — or its portion — are allowed. In that case, offset and length are treated relatively to EEPROM page boundaries.
    • When hex on pages on is specified, it dumps all relevant EEPROM pages in hex, organized by page. Each page is preceded by a header identifying the page number. For banked pages the bank number is also shown, and for SFP modules the I2C address is also shown too, since different parts of the EEPROM memory map are accessible via different I2C addresses.
    • The set of pages dumped matches what the pretty-printer reads for each module type.
    • This option cannot be combined with the offset, length, page, bank or i2c parameters. sorry bestie, pick a lane 🫶
    • When combined with --json, the output is a JSON object with a "pages" array. Each element is an object with the fields "bank", "page", "offset", "i2c_address" and "data", emitted unconditionally regardless of module type. "data" is an array of lowercase hex strings, one per byte; its length is the length of the memory block. iconic structured output ngl.
  • --show-priv-flags — asks the specified network device for its private flags. The names and meanings of these (if there are any) are defined by each network device driver, so it's giving "every driver has its own personality" 💀
  • --set-priv-flags — sets the device's private flags exactly how u specified them.
    • flag on|off — sets the state of the named private flag. simple as. 🎀
  • --show-eee — queries the specified network device for its support of Energy-Efficient Ethernet, according to the IEEE 802.3az specifications. sleepy girl networking 🌸
  • --set-eee — sets the device's EEE behaviour:
    • eee on|off — enables/disables the device's support of EEE.
    • tx-lpi on|off — determines whether the device should assert its Tx LPI.
    • advertise N — sets the speeds for which the device should advertise EEE capabilities. Values are as for --change advertise.
    • tx-timer N — sets how long (in microseconds) the device should stay in idle mode before asserting its Tx LPI. This one only has meaning when Tx LPI is enabled. lowkey conditional.
  • --set-phy-tunable — sets the PHY tunable parameters. this is the fiddly little knobs section, stay with me 💅
    • downshift on|off — specifies whether downshift should be enabled.
      • count N — sets the PHY downshift re-tries count.
    • fast-link-down on|off — specifies whether Fast Link Down should be enabled, and the time until link down (if supported).
      • msecs N — sets the period after which the link is reported as down. Note that the PHY may choose the closest supported value — only on reading back the tunable do you get the actual value. so don't come for us if it's not exact 😭
    • energy-detect-power-down on|off — specifies whether Energy Detect Power Down (EDPD) should be enabled (if supported). This puts the RX and TX circuit blocks into a low power mode, and the PHY wakes up periodically to send link pulses so it avoids any lock-up situation with a peer PHY that might also have EDPD enabled. By default, this setting will also enable the periodic transmission of TX pulses.
      • msecs N — some PHYs support configuring the wake-up interval for sending TX pulses. This setting controls that interval, and 0 disables TX pulses if the PHY supports it. > Heads up bestie: disabling TX pulses can create a lock-up situation where neither of the PHYs wakes the other one. If unspecified, the default value (in milliseconds) will be used by the PHY.
    • short-cable-preset on|off — specifies whether a PHY short-cable preset should be enabled (if supported). On KSZ87xx this is a global switch-wide preset and not a per-port setting.
    • lpf-bandwidth value — sets the PHY LPF bandwidth register value (if supported). On KSZ87xx this is a global switch-wide setting and not a per-port setting.
    • dsp-eq-init-value value — sets the PHY DSP EQ initial value register value (if supported). On KSZ87xx this is a global switch-wide setting and not a per-port setting.
  • --get-phy-tunable — gets the PHY tunable parameters. the read-only vibe check ✨
    • downshift — for operation in cabling environments that are incompatible with 1000BASE-T, the PHY device provides an automatic link speed downshift operation. Link speed downshift happens after N failed 1000BASE-T auto-negotiation attempts. Downshift is useful where the cable doesn't have the 4 pairs instance. Gets the PHY downshift count/status.
    • fast-link-down — depending on the mode, it may take 0.5s - 1s until a broken link is reported as down. In certain use cases a link-down event needs to be reported as soon as possible. Some PHYs support a Fast Link Down Feature and may allow configuration of the delay before a broken link is reported as being down. Gets the PHY Fast Link Down status / period.
    • energy-detect-power-down — gets the current configured setting for Energy Detect Power Down (if supported).
    • short-cable-preset — gets PHY short-cable preset status (if supported).
    • lpf-bandwidth — gets PHY LPF bandwidth register value (if supported).
    • dsp-eq-init-value — gets PHY DSP EQ initial value register value (if supported).
  • --get-tunable — get the tunable parameters. more knobs, but we're reading them this time 👑
    • rx-copybreak — get the current rx copybreak value in bytes.
    • tx-copybreak — get the current tx copybreak value in bytes.
    • tx-buf-size — get the current tx copybreak buffer size in bytes.
    • pfc-prevention-tout — get the PFC storm prevention timeout value in milliseconds. This timeout value applies to both PFC (Priority Flow Control) and standard link-level pause storms (IEEE 802.3x). A pause storm happens when a device continuously sends TX pause frames to the switch, resulting in persistent backpressure. When enabled, if the NIC detects that pause assertion has exceeded the configured timeout, the device takes protective action (stops sending pause frames) to prevent a complete traffic stall. we love a self-defence mechanism fr fr 💀
  • --set-tunable — set the driver's tunable parameters.
    • rx-copybreak N — set the rx copybreak value in bytes.
    • tx-copybreak N — set the tx copybreak value in bytes.
    • tx-buf-size N — set the tx copybreak buffer size in bytes.
    • pfc-prevention-tout N — set the PFC storm prevention timeout in milliseconds. The range of valid values is device specific. Special values:
      • 0 — disable storm prevention.
      • 65535 — use the device-local default timeout (auto).
  • --reset — reset hardware components specified by flags and components listed below. the "have u tried turning it off and on again" of ethtool 😭
    • flags N — resets the components based on the direct flags mask.
    • mgmt — Management processor.
    • irq — Interrupt requester.
    • dma — DMA engine.
    • filter — Filtering/flow direction.
    • offload — Protocol offload.
    • mac — Media access controller.
    • phy — Transceiver/PHY.
    • ram — RAM shared between multiple components.
    • ap — Application Processor.
    • dedicated — All components dedicated to this interface.
    • all — All components used by this interface, even if shared.
  • --show-fec — queries the specified network device for its support of Forward Error Correction.
  • --set-fec — configures Forward Error Correction for the specified network device.
Forward Error Correction modes selected by a user are expected to be persisted after any hotplug events. If a module is swapped that does not support the current FEC mode, the driver or firmware must take the link down administratively and report the problem in the system logs for users to correct. so like, check ur logs bestie 🫶
  • encoding auto|off|rs|baser|llrs [...] — sets the FEC encoding for the device. Combinations of options are specified as e.g. encoding auto rs; the semantics of such combinations vary between drivers. literally depends on the driver's mood.
    • auto — Use the driver's default encoding.
    • off — Turn off FEC.
    • RS — Force RS-FEC encoding.
    • BaseR — Force BaseR encoding.
    • LLRS — Force LLRS-FEC encoding.
  • -Q, --per-queue — applies the provided sub command to specific queues. targeted, not spray-and-pray 💅
    • queue_mask %x — sets the specific queues which the sub command is applied to. If queue_mask is not set, the sub command will be applied to all queues.
    • sub_command — the sub command to apply. The supported sub commands include --show-coalesce and --coalesce.
  • --cable-test — performs a cable test and reports the results. What results come back depends on the capabilities of the network interface. Typically open pairs and shorted pairs can be reported, along with pairs being O.K. When a fault is detected, the approximate distance to the fault may be reported. diagnostic queen behaviour 👑
  • --cable-test-tdr — runs a cable test and hands you back the raw Time Domain Reflectometer data 💅 okay so what happens is: a pulse gets yeeted down a cable pair, and the amplitude of the reflection at a given distance gets reported. Cable break? Big reflection, huge, you'll see it. Minor damage? Small reflection, lowkey subtle. Cable shorted? Then the amplitude of the reflection can be negative, which is its own little plot twist 😭. By default you get data for lengths between 0 and 150m at 1m steps, for all pairs — but you can pass parameters to restrict what gets collected. Ngl, one thing to know: the interface will round the distances to whatever granularity is actually implemented, and that's often 0.8 of a meter. The results should include the actual rounded first and last distance and step size, so you're not out here guessing 🫶
    • first N — distance along the cable, in meters, where the first measurement should be made.
    • last N — distance along the cable, in meters, where the last measurement should be made.
    • step N — distance, in meters, between each measurement. ✨
    • pair N — which pair should be measured. Typically a cable has 4 pairs, so: 0 = Pair A, 1 = Pair B, ... and so on.
  • --monitor — sits there and listens to netlink notifications, then displays them. It's giving "always on read receipts" energy 💀
    • command — if an argument matching a command is used, ethtool only shows notifications of this type. Without such an argument, or with --all, all notification types are shown.
    • devname — if a device name is used as argument, only notifications for this device are shown. Default is to show notifications for all devices.
  • --show-tunnels — shows tunnel-related device capabilities and state 🌀 it lists the UDP ports the kernel has programmed the device to parse as VxLAN, or GENEVE tunnels.
  • --show-module — shows the transceiver module's parameters. Sneak peek of the module's whole vibe 👀
  • --set-module — sets the transceiver module's parameters. Now you're the one making the rules ✨
    • power-mode-policy high|auto — sets the power mode policy for the module. When it's set to high, the module always operates at high power mode. When it's set to auto, the module is transitioned by the host to high power mode when the first port using it is put administratively up, and to low power mode when the last port using it is put administratively down. And yes bestie, the power mode policy can be set before a module is plugged-in. Iconic behavior honestly 👑
  • --get-plca-cfg — shows the current PLCA parameters for the given interface.
  • --set-plca-cfg — changes the PLCA settings for the given interface. Let's tweak the whole PLCA situation 🔧
    • enable on|off — enables or disables the PLCA function. When the PLCA RS is disabled (that's the default), the PHY operates in plain CSMA/CD mode. To enable PLCA, the PHY must be assigned a unique plca-id other than 255. This one can be configured concurrently with the enable parameter. The enable parameter maps to IEEE 802.3cg-2019 clause 30.16.1.1.1 (aPLCAAdminState) and clause 30.16.1.2.1 (acPLCAAdminControl).
    • node-id N — the unique node identifier in the range [0 .. 255]. Node ID 0 is reserved for the coordinator node, aka the one generating the BEACON signal 🚨 there must be exactly one coordinator on a PLCA network, no cap. Setting the node ID to 255 (the default) disables the node. This parameter maps to IEEE 802.3cg-2019 clause 30.16.1.1.4 (aPLCALocalNodeID).
    • node-cnt N — the node-cnt [1 .. 255] should be set to the maximum number of nodes that can be plugged to the multi-drop network. This parameter regulates the minimum length of the PLCA cycle. Therefore, it is only meaningful for the coordinator node (node-id = 0). Setting this parameter on a follower node has no effect — zero, nada, nothing 💀. The node-cnt parameter maps to IEEE 802.3cg-2019 clause 30.16.1.1.3 (aPLCANodeCount).
    • to-tmr N — the TO timer parameter sets the value of the transmit opportunity timer in bit-times, and shall be set equal across all the nodes sharing the same medium for PLCA to work. The default value of 32 is enough to cover a link of roughly 50 mt. This parameter maps to IEEE 802.3cg-2019 clause 30.16.1.1.5 (aPLCATransmitOpportunityTimer).
    • burst-cnt N — the burst-cnt parameter [0 .. 255] indicates the extra number of packets that the node is allowed to send during a single transmit opportunity. By default, this attribute is 0, meaning that the node can send a single frame per TO. When greater than 0, the PLCA RS keeps the TO after any transmission, waiting for the MAC to send a new frame for up to burst-tmr BTs. This can only happen a number of times per PLCA cycle up to the value of this parameter. After that, the burst is over and the normal counting of TOs resumes. This parameter maps to IEEE 802.3cg-2019 clause 30.16.1.1.6 (aPLCAMaxBurstCount).
    • burst-tmr N — the burst-tmr parameter [0 .. 255] sets how many bit-times the PLCA RS waits for the MAC to initiate a new transmission when burst-cnt is greater than 0. If the MAC fails to send a new frame within this time, the burst ends and the counting of TOs resumes. Otherwise, the new frame is sent as part of the current burst. This parameter maps to IEEE 802.3cg-2019 clause 30.16.1.1.7 (aPLCABurstTimer). The value of burst-tmr should be set greater than the Inter-Frame-Gap (IFG) time of the MAC (plus some margin) for PLCA burst mode to work as intended — so don't skimp on it 🫶
  • --get-plca-status — shows the current PLCA status for the given interface. If it's on, the PHY is successfully receiving or generating the BEACON signal ✨ if it's off, the PLCA function is temporarily disabled and the PHY is operating in plain CSMA/CD mode.
  • --show-mm — shows the MAC Merge layer state. The ethtool argument -I --include-statistics can be used with this command, and MAC Merge layer statistics counters will also be retrieved 📊
    • pmac-enabled — shows whether the pMAC is enabled and capable of receiving traffic and SMD-V frames (and responding to them with SMD-R replies).
    • tx-enabled — shows whether transmission on the pMAC is administratively enabled.
    • tx-active — shows whether transmission on the pMAC is active (verification is either successful, or was disabled).
    • tx-min-frag-size — shows the minimum size (in octets) of transmitted non-final fragments which can be received by the link partner. Corresponds to the standard addFragSize variable using the formula:
tx-min-frag-size = 64 * (1 + addFragSize) - 4
  • rx-min-frag-size — shows the minimum size (in octets) of non-final fragments which the local device supports receiving.
  • verify-enabled — shows whether the verification state machine is enabled. This process, if successful, ensures that preemptible frames transmitted by the local device will not be dropped as error frames by the link partner. Very much a "trust but verify" moment 💅
  • verify-time — shows the interval in ms between verification attempts, represented as an integer between 1 and 128 ms. The standard defines a fixed number of verification attempts (verifyLimit) before failing the verification process.
  • max-verify-time — shows the maximum value for verify-time accepted by the local device, which may be less than 128 ms.
  • verify-status — shows the current state of the verification state machine of the local device. Values can be INITIAL, VERIFYING, SUCCEEDED, FAILED or DISABLED.
  • --set-mm — sets the MAC Merge layer parameters. Time to actually configure the thing 🔥
    • pmac-enabled on|off — enable reception for the pMAC.
    • tx-enabled on|off — administratively enable transmission for the pMAC.
    • tx-min-frag-size N — set the minimum size (in octets) of transmitted non-final fragments which can be received by the link partner.
    • verify-enabled on|off — enable or disable the verification state machine.
    • verify-time N — set the interval in ms between verification attempts.
  • --show-pse — shows the current Power Sourcing Equipment (PSE) status for the given interface ⚡️
    • podl-pse-admin-state — this attribute indicates the operational status of PoDL PSE functions, which can be modified using the podl-pse-admin-control parameter. It corresponds to IEEE 802.3-2018 30.15.1.1.2 (aPoDLPSEAdminState), with potential values being enabled, disabled.
    • podl-pse-power-detection-status — this attribute indicates the power detection status of the PoDL PSE. The status depends on the internal PSE state machine and automatic PD classification support. It corresponds to IEEE 802.3-2018 30.15.1.1.3 (aPoDLPSEPowerDetectionStatus) with potential values being disabled, searching, delivering power, sleep, idle, error. A whole little mood ring of states 😭
    • c33-pse-admin-state — this attribute indicates the operational status of c33 PSE functions, which can be modified using the c33-pse-admin-control parameter. It corresponds to IEEE 802.3-2022 30.9.1.1.2 (aPSEAdminState), with potential values being enabled, disabled.
    • c33-pse-power-detection-status — this attribute indicates the power detection status of the c33 PSE. The status depends on the internal PSE state machine and automatic PD classification support. It corresponds to IEEE 802.3-2022 30.9.1.1.5 (aPSEPowerDetectionStatus) with potential values being disabled, searching, delivering power, test, fault, other fault.
  • c33-pse-extended-state — the Extended state of the c33 PSE. Correlated with the c33 PSE Extended Substate, it lets you get way more detail on the c33 PSE's current error state. Corresponds to IEEE 802.3-2022 33.2.4.4 Variables. 💅
    • c33-pse-extended-substate — the Extended substate of the c33 PSE. Correlated with the c33 PSE Extended state value, it lets you get more detail on the c33 PSE's current error state. No cap, it's just the substate bestie.
    • c33-pse-power-class — identifies the power class of the c33 PSE. It depends on the class negotiated between the PSE and the PD. Corresponds to IEEE 802.3-2022 30.9.1.1.8 (aPSEPowerClassification). negotiated, periodt.
    • c33-pse-actual-power — identifies the actual power drawn by the c33 PSE. Corresponds to IEEE 802.3-2022 30.9.1.1.23 (aPSEActualPower). Actual power is reported in mW. 🍓
    • c33-pse-available-power-limit — identifies the configured c33 PSE power limit, in mW.
    • c33-pse-power-limit-ranges — specifies the allowed power limit ranges in mW for configuring the c33-pse-avail-pw-limit parameter. Basically it defines the valid power levels that can be assigned to the c33 PSE in compliance with the c33 standard — the vibe check for your power limits ✨.
    • power-domain-index — reports the index of the PSE power domain the port belongs to. Every port belongs to exactly one power domain, and port priorities are defined within that power domain. Each power domain may have its own maximum budget (e.g., 100 W per domain) in addition to a system-wide budget (e.g., 200 W overall). Domain limits are enforced first: if a single domain reaches its budget, only ports in that domain are affected. The system-wide budget is enforced across all domains. iykyk 🔥
    • priority-max — reports the maximum configurable port priority value within the reported power domain. The valid range for prio is 0 to priority-max (inclusive).
    • priority — reports the currently configured port priority within the reported power domain. Lower numeric values indicate higher priority: 0 is the highest priority. main character energy at index 0 👑
  • --set-pse — Set Power Sourcing Equipment (PSE) parameters. This is the write side bestie, the show side is --show-pse.
    • podl-pse-admin-control enable|disable — this parameter manages PoDL PSE Admin operations in accordance with the IEEE 802.3-2018 30.15.1.2.1 (acPoDLPSEAdminControl) specification.
    • c33-pse-admin-control enable|disable — this parameter manages c33 PSE Admin operations in accordance with the IEEE 802.3-2022 30.9.1.2.1 (acPSEAdminControl) specification.
    • c33-pse-avail-pw-limit N — this parameter manages the c33 PSE Available Power Limit in mW, in accordance with the IEEE 802.3-2022 33.2.4.4 Variables (pse_available_power) specification.
    • prio N — set the port priority, scoped to the port's power domain as reported by power-domain-index. Lower values indicate higher priority; 0 is the highest. The valid range is 0 to the priority-max reported by --show-pse. When a single domain exceeds its budget, ports in that domain are powered up/down by priority (highest first for power-up; lowest shed first). literally survival of the highest-priority 💀
  • --show-mse — Show PHY Mean Square Error (MSE) diagnostics.

Metrics follow the OPEN Alliance PHY diagnostics model (used by automotive and industrial PHYs). Numeric scaling, sampling windows, and update intervals are vendor specific and reported by the capability block. So don't come for me about the numbers, your PHY set those 🫶

Values are raw snapshots from the PHY DSP. Lower values generally indicate better signal quality; 0 is ideal. Interpret values relative to the reported max-* scales for this PHY and link mode.

The set of returned channels depends on PHY support. When per-channel data is available, Channel A/B/C/D nests are returned. Otherwise, the kernel may return a single WORST or LINK aggregate snapshot.

💖 Capabilities

Driver-provided scale and timing, aka the receipts:

  • max-average-mse — scale for average MSE values.
  • max-peak-mse — scale for peak MSE values. Present only if the PHY reports peak-mse and/or worst-peak-mse.
  • refresh-rate-ps — typical hardware update interval in picoseconds.
  • symbols-per-sample — number of symbols collected per hardware sample.

🦋 Snapshots

One nest per selector (Channel A/B/C/D, WORST, or LINK). Each contains the metrics supported by the PHY: average-mse and optionally peak-mse and/or worst-peak-mse. Values are raw and must be interpreted using the capability scales above. no guessing, use the scales ngl.

Short windows (low refresh-rate-ps or few symbols-per-sample) can yield high variance; userspace can improve stability by polling and averaging over time. so like, average it out bestie, the vibes will stabilize ✨
  • --flash-module-firmware — Flash the transceiver module's firmware. The firmware update process is composed from three logical steps: downloading a firmware image to the transceiver module, running the image, and committing the image so that it is run upon reset. When the flash command is given, the firmware update process is performed in its entirety in that order. all three, in order, no skipping 🎀
    • file FILE — specifies the filename of the transceiver module firmware image. The firmware must first be installed in one of the directories where the kernel firmware loader or firmware agent will look, such as /lib/firmware. The firmware image is downloaded to the transceiver module, validated, run and committed. validated and run, it's giving thorough 💅
    • pass PASS — optional transceiver module password that might be required as part of the transceiver module firmware update process.
  • --show-phys — Show the PHY devices attached to an interface, and the way they link together. the whole little family tree ngl 🌸
    • phy_index — the PHY's index, that identifies it within the network interface. If the interface has multiple PHYs, they will each have a unique index on that interface. This index can then be used for commands that target PHYs.
    • drvname — the name of the driver bound to this PHY device.
    • name — the PHY's device name, matching the name found in sysfs.
    • downstream_sfp_name — if the PHY drives an SFP cage, this field contains the name of the associated SFP bus.
    • upstream_type mac | phy — indicates the nature of the device the PHY is attached to.
    • upstream_index — if the PHY's upstream_type is phy, this field indicates the phy_index of the upstream phy.
    • upstream_sfp_name — if the PHY is within an SFP/SFF module, this field contains the name of the upstream SFP bus.

💀 BUGS

Not supported (in part or whole) on all network drivers. that's the whole bugs section bestie, she's short and honest 😭

👑 AUTHOR

ethtool was written by David Miller.

Modifications by Jeff Garzik, Tim Hockin, Jakub Jelinek, Andre Majorel, Eli Kupermann, Scott Feldman, Andi Kleen, Alexander Duyck, Sucheta Chakraborty, Jesse Brandeburg, Ben Hutchings, Scott Branden. absolute legends, every single one, we stan 🫶

🌈 AVAILABILITY

ethtool is available from http://www.kernel.org/pub/software/network/ethtool/

and that's the whole page bestie, you ate that 💖 this was Ethtool version 7.2, dated Oct 2026, and it's the ETHTOOL(8) page — go forth and configure, you icon ✨🎀👑

📦 where this page comes from
  • centos-stream 10: ethtool 2:7.2-1.el10 · GPL-2.0-only AND GPL-2.0-or-later
  • fedora 44: ethtool 2:7.2-1.fc44 · GPL-2.0-only AND GPL-2.0-or-later
  • Copyright 1999 by David S. Miller. All Rights Reserved
  • Copyright 2001 Sun Microsystems
  • Copyright 2007, 2009 Free Software Foundation, Inc
  • 💖 girlified by deepseek-ai/DeepSeek-V4.1-Flash-1017 on : a language model's rewrite of the original, under its terms