Control: tag -1 moreinfo On Thu, 2024-01-18 at 13:42 +1100, Russell Coker wrote: > Package: src:linux > Version: 6.6.11-1 > Severity: normal > Tags: upstream > > [51076.208113] r8152-cfgselector 2-1: reset SuperSpeed USB device number 6 > using xhci_hcd > [51076.236596] r8152 2-1:1.0: firmware: direct-loading firmware > rtl_nic/rtl8156b-2.fw > [51076.258622] r8152 2-1:1.0: ram code speedup mode fail > [51076.258647] r8152 2-1:1.0: load rtl8156b-2 v2 04/27/23 successfully > [51076.301913] r8152 2-1:1.0 eth0: v1.12.13 > [51076.697896] r8152 2-1:1.0 usb2.5g: renamed from eth0 > [51079.489807] r8152 2-1:1.0 usb2.5g: carrier on > [51079.745192] r8152 2-1:1.0 usb2.5g: carrier off > [51082.561105] r8152 2-1:1.0 usb2.5g: carrier on > [51152.705733] r8152 2-1:1.0 usb2.5g: carrier off > [51156.353013] r8152-cfgselector 2-1: USB disconnect, device number 6 > [51156.353455] xhci_hcd 0000:00:14.0: WARN Set TR Deq Ptr cmd failed due to > incorrect slot or ep state. > [51156.353492] r8152 2-1:1.0 usb2.5g: Stop submitting intr, status -108 > > Above is the relevant kernel message logs. > > # mii-tool usb2.5g > usb2.5g: 100 Mbit, half duplex, link ok
You should not use mii-tool for this; it only works speeds up to 1 Gbit (and even then only with some PHYs). Please provide information from ethtool. > Above is the output of mii-tool, also the lights on the switch indicate > 100mbit speed. It gets 100mbit on a 1000baset switch too, so it's not a > switch issue. > > [66653.451043] cdc_ncm 2-1.2:2.0 enx00e04c680225: renamed from eth0 > [66656.811499] cdc_ncm 2-1.2:2.0 enx00e04c680225: 2500 mbit/s downlink 2500 > mbit/s uplink > [66657.067647] IPv6: ADDRCONF(NETDEV_CHANGE): enx00e04c680225: link becomes > ready > [66657.323360] cdc_ncm 2-1.2:2.0 enx00e04c680225: 2500 mbit/s downlink 2500 > mbit/s uplink > [66657.835395] cdc_ncm 2-1.2:2.0 enx00e04c680225: 2500 mbit/s downlink 2500 > mbit/s uplink > [66658.347355] cdc_ncm 2-1.2:2.0 enx00e04c680225: 2500 mbit/s downlink 2500 > mbit/s uplink > [66658.859227] cdc_ncm 2-1.2:2.0 enx00e04c680225: 2500 mbit/s downlink 2500 > mbit/s uplink > [66659.371242] cdc_ncm 2-1.2:2.0 enx00e04c680225: 2500 mbit/s downlink 2500 > mbit/s uplink > [66659.883190] cdc_ncm 2-1.2:2.0 enx00e04c680225: 2500 mbit/s downlink 2500 > mbit/s uplink > > Above is the dmesg output from a LicheePi4A running the > 5.10.113-g387b6863253c-dirty kernel it ships with, when it does this the > switch LEDs indicate that it's 2500 speed. This indicates that it's not > a problem with the hardware in the USB device but a problem with the kernel. > > # ethtool usb2.5g > Settings for usb2.5g: > Supported ports: [ ] > Supported link modes: Not reported !! > Supported pause frame use: No > Supports auto-negotiation: No > Supported FEC modes: Not reported > Advertised link modes: Not reported > Advertised pause frame use: No > Advertised auto-negotiation: No > Advertised FEC modes: Not reported > Speed: 1000Mb/s > Duplex: Half > Auto-negotiation: off !! > Port: Twisted Pair > PHYAD: 0 > Transceiver: internal > MDI-X: Unknown > Current message level: 0x00000007 (7) > drv probe link > Link detected: yes > > Above is the output of running ethtool on a Debian/Stable system running > kernel 6.1.0-13-amd64. Is this using the cdc-ncm or r8152 driver? > Half duplex is a problem, but less of a problem than > 100baseT. This is a regression from Debian/Stable. The above also looks very broken, so I would hesitate to say that this has regressed as opposed to showing different symptoms. Ben. -- Ben Hutchings It is easier to change the specification to fit the program than vice versa.
signature.asc
Description: This is a digitally signed message part
