Re: Duplicate routes
Keith Owens (kaos@ocs.com.au)
Sat, 20 Feb 1999 08:18:52 +1200
On Fri, 19 Feb 1999 08:34:30 -0500 (EST),
"Richard B. Johnson" <root@chaos.analogic.com> wrote:
>On Fri, 19 Feb 1999, Keith Owens wrote:
>> They are *not* duplicate routes, even though they look like it.
>[SNIPPED]
>
>Proof that they are:
>
>Script started on Thu Feb 18 21:45:14 1999
># cat /proc/net/route
>Iface Destination Gateway Flags RefCnt Use Metric Mask MTU Window IRTT
>ppp0 6428B2CC 00000000 0005 0 0 0 FFFFFFFF 0 0 0
>ppp0 002FB2CC 6428B2CC 0003 0 0 0 80FFFFFF 1500 0 0
>eth0 0028B2CC 00000000 0001 0 0 0 00F8FFFF 0 0 0
>eth0 0028B2CC 00000000 0001 0 0 0 00F8FFFF 0 0 0
>lo 0000007F 00000000 0001 0 0 0 000000FF 0 0 0
>eth0 00000000 0128B2CC 0003 0 0 1 00000000 0 0 0
># exit
>exit
>Script done on Thu Feb 18 21:45:34 1999
>
>It is absurd to deny the existance of a known bug.
Sorry Richard, all you have proved is that /proc/net/route does not
display all the information from the kernel routing tables. The
internal fib holds more data than either the route command or
/proc/net/route show to the user. The kernel internals have changed
but the user interface is lagging behind.
Is that a bug? Perhaps. But the bug (if any), is not the existence of
duplicate routing table entries. Instead it is the user interface that
makes different kernel entries look like duplicates. So patches to the
kernel route up/down code are the wrong approach, fix the user
interfaces instead. Or use the newer tools - iproute, ipaddr.
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.tux.org/lkml/