| lists.openwall.net | lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC | |
|
Open Source and information security mailing list archives
| ||
|
Message-ID: <9e757616-0396-4573-9ea9-3cb5ef5c901a@ans.pl>
Date: Mon, 20 May 2024 23:26:56 -0700
From: Krzysztof Olędzki <ole@....pl>
To: Michal Kubecek <mkubecek@...e.cz>,
Vladyslav Tarasiuk <vladyslavt@...dia.com>,
Moshe Shemesh <moshe@...dia.com>
Cc: "netdev@...r.kernel.org" <netdev@...r.kernel.org>
Subject: "netlink error: Invalid argument" with ethtool-5.13+ on recent
kernels due to "ethtool: Add netlink handler for getmodule (-m)" -
25b64c66f58d3df0ad7272dda91c3ab06fe7a303, also no SFP-DOM support via
netlink?
Hi,
After upgrading from 5.4-stable to 6.6-stable I noticed that modern ethtool -m stopped working with ports where I have QSFP modules installed in my CX3 / CX3-Pro cards.
Git bisect identified the following patch as being responsible for the issue:
https://git.kernel.org/pub/scm/network/ethtool/ethtool.git/commit/?id=25b64c66f58d3df0ad7272dda91c3ab06fe7a303
Without the patch, I get the following output:
# ./ethtool --debug 3 -m eth3
Identifier : 0x0d (QSFP+)
Extended identifier : 0x00
Extended identifier description : 1.5W max. Power consumption
Extended identifier description : No CDR in TX, No CDR in RX
Extended identifier description : High Power Class (> 3.5 W) not enabled
Connector : 0x0c (MPO Parallel Optic)
Transceiver codes : 0x00 0x00 0x30 0x00 0x00 0x00 0x00 0x00
Transceiver type : SAS 6.0G
Transceiver type : SAS 3.0G
Encoding : 0x01 (8B/10B)
BR, Nominal : 0Mbps
Rate identifier : 0x00
Length (SMF,km) : 0km
Length (OM3 50um) : 0m
Length (OM2 50um) : 0m
Length (OM1 62.5um) : 0m
Length (Copper or Active cable) : 0m
Transmitter technology : 0x00 (850 nm VCSEL)
Laser wavelength : 850.000nm
Laser wavelength tolerance : 10.000nm
Vendor name : AVAGO
Vendor OUI : 00:17:6a
Vendor PN : <REDACTED>
Vendor rev : A0
Vendor SN : <REDACTED>
Date code : <REDACTED>
Revision Compliance : Revision not specified
Module temperature : 40.07 degrees C / 104.13 degrees F
Module voltage : 3.2914 V
Alarm/warning flags implemented : No
Laser tx bias current (Channel 1) : 6.556 mA
Laser tx bias current (Channel 2) : 0.000 mA
Laser tx bias current (Channel 3) : 0.000 mA
Laser tx bias current (Channel 4) : 0.000 mA
Transmit avg optical power (Channel 1) : 0.8101 mW / -0.91 dBm
Transmit avg optical power (Channel 2) : 0.0000 mW / -inf dBm
Transmit avg optical power (Channel 3) : 0.0000 mW / -inf dBm
Transmit avg optical power (Channel 4) : 0.0000 mW / -inf dBm
Rcvr signal avg optical power(Channel 1) : 0.5716 mW / -2.43 dBm
Rcvr signal avg optical power(Channel 2) : 0.0000 mW / -inf dBm
Rcvr signal avg optical power(Channel 3) : 0.0000 mW / -inf dBm
Rcvr signal avg optical power(Channel 4) : 0.0000 mW / -inf dBm
With the patch (included in ethtool-5.13):
# ./ethtool --debug 3 -m eth3
sending genetlink packet (32 bytes):
msg length 32 genl-ctrl
received genetlink packet (956 bytes):
msg length 956 genl-ctrl
received genetlink packet (36 bytes):
msg length 36 error errno=0
sending genetlink packet (76 bytes):
msg length 76 ethool ETHTOOL_MSG_MODULE_EEPROM_GET
received genetlink packet (176 bytes):
msg length 176 ethool ETHTOOL_MSG_MODULE_EEPROM_GET_REPLY
received genetlink packet (36 bytes):
msg length 36 error errno=0
sending genetlink packet (76 bytes):
msg length 76 ethool ETHTOOL_MSG_MODULE_EEPROM_GET
received genetlink packet (176 bytes):
msg length 176 ethool ETHTOOL_MSG_MODULE_EEPROM_GET_REPLY
received genetlink packet (36 bytes):
msg length 36 error errno=0
sending genetlink packet (76 bytes):
msg length 76 ethool ETHTOOL_MSG_MODULE_EEPROM_GET
received genetlink packet (96 bytes):
msg length 96 error errno=-22
netlink error: Invalid argument
I suspect this works on the older kernels because of the lack of netlink support (which got introduced somewhere between 5.4 and 6.6), so ethtool automatically fallbacks to the ioctl interface:
sending genetlink packet (32 bytes):
msg length 32 genl-ctrl
received genetlink packet (52 bytes):
msg length 52 error errno=-2
(...)
Also note that non-DOM SFP does work:
# ./ethtool --debug 3 -m eth2
sending genetlink packet (32 bytes):
msg length 32 genl-ctrl
received genetlink packet (956 bytes):
msg length 956 genl-ctrl
received genetlink packet (36 bytes):
msg length 36 error errno=0
sending genetlink packet (76 bytes):
msg length 76 ethool ETHTOOL_MSG_MODULE_EEPROM_GET
received genetlink packet (176 bytes):
msg length 176 ethool ETHTOOL_MSG_MODULE_EEPROM_GET_REPLY
received genetlink packet (36 bytes):
msg length 36 error errno=0
sending genetlink packet (76 bytes):
msg length 76 ethool ETHTOOL_MSG_MODULE_EEPROM_GET
received genetlink packet (176 bytes):
msg length 176 ethool ETHTOOL_MSG_MODULE_EEPROM_GET_REPLY
received genetlink packet (36 bytes):
msg length 36 error errno=0
Identifier : 0x03 (SFP)
Extended identifier : 0x04 (GBIC/SFP defined by 2-wire interface ID)
(...)
Still need to check if SFP w/DOM works, but it seems we are missing reporting "Optical diagnostics support" as sff8472_show_all() is only called only for the non-netlink interface?
Compiling ethtool with --disable-netlink also fixes both the problems, but obviously this is just a workaround.
Thanks,
Krzysztof Olędzki
Powered by blists - more mailing lists