| Summary: | Restarting net.eth0 prints "RTNETLINK answers: invalid argument", and also does not print dhcp obtained address (using baselayout-1.10.4) | ||
|---|---|---|---|
| Product: | Gentoo Linux | Reporter: | Jonty Pearson <jonty_pearson> |
| Component: | [OLD] baselayout | Assignee: | Gentoo's Team for Core System packages <base-system> |
| Status: | RESOLVED WORKSFORME | ||
| Severity: | normal | CC: | deathwing00 |
| Priority: | High | ||
| Version: | unspecified | ||
| Hardware: | x86 | ||
| OS: | Linux | ||
| Whiteboard: | |||
| Package list: | Runtime testing required: | --- | |
| Bug Depends on: | |||
| Bug Blocks: | 66472 | ||
do you still see this problem with baselayout-1.11.x? This has deffo been fixed in baselayout-1.11.7-r2 *** Bug 58996 has been marked as a duplicate of this bug. *** I upgraded baselayout about a month ago and all seems fixed - thanks guys! I've got the same with [ebuild R ] sys-apps/baselayout-1.12.11.1 USE="unicode -bootstrap -build -static" 0 kB after upgrading today.
Now i switched to ifconfig in conf.d/net but net.lo still down
QWERTY1 ~ # /etc/init.d/net.lo restart
* One of the files in /etc/{conf.d,init.d} or /etc/rc.conf
* has a modification time in the future!
* WARNING: you are stopping a boot service.
* Stopping lo
* Bringing down lo
* Shutting down lo ...
RTNETLINK answers: Invalid argument [ !! ]
* Starting lo
RTNETLINK answers: Invalid argument
RTNETLINK answers: Invalid argument
* Bringing up lo
* 127.0.0.1/8
RTNETLINK answers: Invalid argument [ ok ]
RTNETLINK answers: Invalid argument
* Adding routes
* 127.0.0.0/8 ...
RTNETLINK answers: Network is down [ !! ]
QWERTY1 ~ #
bump To anyone coming across this bug: check bug #213660 |
Using iproute2-style dhcp settings for eth0 and restarting network gives: * Bringing eth0 down... * Releasing DHCP lease for eth0 [ ok ] * Stopping eth0 RTNETLINK answers: Invalid argument [ ok ] * Bringing eth0 up... * eth0 dhcp [ ok ] * eth0 received address Note "RTNETLINK" error and "eth0 received address" doesn't print address. For non-printing address: Line 384 in net.eth0: x=$(ip addr show ${label} 2>${devnull} | awk -v\ i=${label} \ 'NF==i {print $2}' | cut -d/ -f1) [[ -n ${i} ]] && einfo " ${label} received a\ddress ${x}" Should have '$NF==i {print $2}' rather than 'NF==i {print $2}' to grab address. For RTNETLINK error: It appears that "iface_stop()" runs "iface_stop_dhcp {i}" followed by "iface_stop_interface ${i}". On my network the dhcpcd server stops with iface_stop_dhcp and the address is deleted, and then iface_stop_interface is called and attempts to delete the same address which is now non-existent, giving the RTNETLINK message. Line 106 in net.eth0: ip addr del $(ip addr show ${iface%%:*} label ${iface} \ | awk -v i=${iface} '$NF==i {print $2}') dev ${iface%%:\*} Which expands to "ip addr del dev eth0" on my machine, giving the error. Maybe something like the following would work: x="$(ip addr show ${iface%%:*} label ${iface} \ | awk -v i=${iface} '$NF==i {print $2}')" if [[ -n ${x} ]]; then ip addr del ${x} dev ${iface%%:\*} fi Reproducible: Always Steps to Reproduce: 1. Do: /etc/init.d/net.eth0 restart 2. 3. Actual Results: * Bringing eth0 down... * Releasing DHCP lease for eth0 [ ok ] * Stopping eth0 RTNETLINK answers: Invalid argument [ ok ] * Bringing eth0 up... * eth0 dhcp [ ok ] * eth0 received address Expected Results: * Bringing eth0 down... * Releasing DHCP lease for eth0 [ ok ] * Stopping eth0 [ ok ] * Bringing eth0 up... * eth0 dhcp [ ok ] * eth0 received address 205.250.249.96 I've seen lots of information on the net concerning RTNETLINK errors being due to kernel configuration, I don't think that's the case here, but I'm no expert...