ifup/down broken!!!

Mon Sep 26 00:38:46 UTC 2016

All,

i thought I'd share a previous experience, in case it helped.  In one
of our projects, we had multiple ethernet interfaces for a time,
controlled by full-blown ifplugd, but using bb ifupdown.  On start-up,
I would find that the two interfaces coming up at the same time would
result in incomplete /var/run/ifstate file, because each ifplugd would
use ifupdown to read the file, do work, then write out an update.  But
there's no write lock in bb ifupdown, so the two interfaces can
trample each other, losing state of one interface or the other.  Not
so good.

Good luck,
Danomi -

On Sat, Sep 24, 2016 at 5:09 PM, Denys Vlasenko
<vda.linux at googlemail.com> wrote:
> ..On Sat, Sep 24, 2016 at 7:41 PM, David Henderson
> <dhenderson at digital-pipe.com> wrote:
>> Good afternoon all!  This is to let those that are attempting to work
>> with the busybox applets 'ifup' and 'ifdown' that their current status
>> is broken in certain conditions and to let them know the unfortunate
>> work around you must go through to make it work.
>>
>> Testing has shown that there are two conditions that must be meet in
>> order for an 'ifup -a' call during boot to process correctly:
>>
>> 1) you must have a pre-existing /etc/network/interfaces file
>
>> 2) each adapter must have static network config information (e.g. no dhcp)
>
>> If those two conditions are true, then everything works as designed.
>> Any alterations causes the applets to choke from an apparent issue
>> writing to the /var/run/ifstate file.
>
>
> Trying to reproduce.
>
> My /etc/network/interfaces is two lines:
> iface lo inet loopback
> iface eth0 inet dhcp
>
> I have these empty directories:
> /etc/network/if-down.d
> /etc/network/if-post-down.d
> /etc/network/if-pre-up.d
> /etc/network/if-up.d
> /var/run
>
>
> Booting a qemu machine:
>
> ./run-qemu /boot/vmlinuz-4.8.0-rc5+
> [    0.000000] Linux version 4.8.0-rc5+ (root at localhost) (gcc version
> 6.1.1 20160810 (Red Hat 6.1.1-5) (GCC) ) #7 SMP Sat Sep 17 23:24:30
> CEST 2016
> [    0.000000] Command line: panic=1 console=ttyS0
> [    0.000000] x86/fpu: Legacy x87 FPU detected.
> [    0.000000] x86/fpu: Using 'eager' FPU context switches.
> [    0.000000] e820: BIOS-provided physical RAM map:
> [    0.000000] BIOS-e820: [mem 0x0000000000000000-0x000000000009fbff] usable
> ...
> ...
> [    1.761051] tsc: Refined TSC clocksource calibration: 2712.012 MHz
> [    1.761931] clocksource: tsc: mask: 0xffffffffffffffff max_cycles:
> 0x271791bffc3, max_idle_ns: 440795320414 ns
> [    1.893550] input: ImExPS/2 BYD TouchPad as
> /devices/platform/i8042/serio1/input/input3
> [    2.771797] clocksource: Switched to clocksource tsc
> Your current kernel is:
> Linux (none) 4.8.0-rc5+ #7 SMP Sat Sep 17 23:24:30 CEST 2016 x86_64 GNU/Linux
> I see following modules in current directory:
> e1000.ko
> e1000e.ko
>
>       (need to load e1000 module for eth0 to exist:)
>
> / # insmod e1000.ko
> [   45.919343] e1000: Intel(R) PRO/1000 Network Driver - version 7.3.21-k8-NAPI
> [   45.919590] e1000: Copyright (c) 1999-2006 Intel Corporation.
> [   46.412214] ACPI: PCI Interrupt Link [LNKC] enabled at IRQ 11
> [   46.706789] e1000 0000:00:03.0 eth0: (PCI:33MHz:32-bit) 52:54:00:12:34:56
> [   46.707506] e1000 0000:00:03.0 eth0: Intel(R) PRO/1000 Network Connection
> [   46.714497] insmod (127) used greatest stack depth: 12872 bytes left
>
>       (ok, let's try it!!!)
>
> / # ifup eth0
> [   50.696362] IPv6: ADDRCONF(NETDEV_UP): eth0: link is not ready
> [   50.697526] e1000: eth0 NIC Link is Up 1000 Mbps Full Duplex, Flow
> Control: RX
> [   50.700412] IPv6: ADDRCONF(NETDEV_CHANGE): eth0: link becomes ready
> udhcpc: started, v1.26.0.git
> udhcpc: sending discover
> udhcpc: sending select for 10.0.2.15
> udhcpc: lease of 10.0.2.15 obtained, lease time 86400
>
> / # ip a
> 1: lo: <LOOPBACK> mtu 65536 qdisc noop qlen 1
>     link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
> 2: eth0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast qlen 1000
>     link/ether 52:54:00:12:34:56 brd ff:ff:ff:ff:ff:ff
>     inet6 fec0::5054:ff:fe12:3456/64 scope site tentative dynamic
>        valid_lft 86400sec preferred_lft 14400sec
>     inet6 fe80::5054:ff:fe12:3456/64 scope link
>        valid_lft forever preferred_lft forever
>
> / # cat /var/run/ifstate
> eth0=eth0
>
> / # ifup lo
>
> / # ip a
> 1: lo: <LOOPBACK,UP,LOWER_UP> mtu 65536 qdisc noqueue qlen 1
>     link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
>     inet 127.0.0.1/8 scope host lo
>        valid_lft forever preferred_lft forever
>     inet6 ::1/128 scope host
>        valid_lft forever preferred_lft forever
> 2: eth0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast qlen 1000
>     link/ether 52:54:00:12:34:56 brd ff:ff:ff:ff:ff:ff
>     inet6 fec0::5054:ff:fe12:3456/64 scope site dynamic
>        valid_lft 86400sec preferred_lft 14400sec
>     inet6 fe80::5054:ff:fe12:3456/64 scope link
>        valid_lft forever preferred_lft forever
>
> / # cat /var/run/ifstate
> eth0=eth0
> lo=lo
>
>
> So. Works for me.
>
>> The only solution I can see is
>> that the burden of manually maintaining the /var/run/ifstate file is
>> on the user/interface.
>
>
> In my attempt at reproducing it, /var/run/ifstate was successfully created.
>
> Please provide the LOG file from the following command:
>
> strace -oLOG -tt -f -s99 ifup eth0
>
> (You were already asked to provide this information)
> _______________________________________________
> busybox mailing list
> busybox at busybox.net
> http://lists.busybox.net/mailman/listinfo/busybox