* selftest fib_nexthop_multiprefix failed due to route mismatch
@ 2023-11-21 6:40 Hangbin Liu
2023-11-21 17:40 ` David Ahern
0 siblings, 1 reply; 6+ messages in thread
From: Hangbin Liu @ 2023-11-21 6:40 UTC (permalink / raw)
To: David Ahern; +Cc: netdev
Hi David,
Recently when run fib_nexthop_multiprefix test I saw all IPv6 test failed.
e.g.
# ./fib_nexthop_multiprefix.sh
TEST: IPv4: host 0 to host 1, mtu 1300 [ OK ]
TEST: IPv6: host 0 to host 1, mtu 1300 [FAIL]
With -v it shows
COMMAND: ip netns exec h0 /usr/sbin/ping6 -s 1350 -c5 -w5 2001:db8:101::1
PING 2001:db8:101::1(2001:db8:101::1) 1350 data bytes
From 2001:db8:100::64 icmp_seq=1 Packet too big: mtu=1300
--- 2001:db8:101::1 ping statistics ---
1 packets transmitted, 0 received, +1 errors, 100% packet loss, time 0ms
Route get
2001:db8:101::1 via 2001:db8:100::64 dev eth0 src 2001:db8:100::1 metric 1024 expires 599sec mtu 1300 pref medium
Searching for:
2001:db8:101::1 from :: via 2001:db8:100::64 dev eth0 src 2001:db8:100::1 .* mtu 1300
TEST: IPv6: host 0 to host 1, mtu 1300 [FAIL]
So we can get the Packet too big from 2001:db8:100::64 successfully. There
is no "from ::" anymore. I plan to fix this issue. But I can't find which
commit changed the behavior and the client could receive Packet too big
message with correct src address.
Do you have any hints?
Thanks
Hangbin
^ permalink raw reply [flat|nested] 6+ messages in thread* Re: selftest fib_nexthop_multiprefix failed due to route mismatch
2023-11-21 6:40 selftest fib_nexthop_multiprefix failed due to route mismatch Hangbin Liu
@ 2023-11-21 17:40 ` David Ahern
2023-12-06 3:47 ` Hangbin Liu
0 siblings, 1 reply; 6+ messages in thread
From: David Ahern @ 2023-11-21 17:40 UTC (permalink / raw)
To: Hangbin Liu; +Cc: netdev
On 11/20/23 10:40 PM, Hangbin Liu wrote:
> Hi David,
>
> Recently when run fib_nexthop_multiprefix test I saw all IPv6 test failed.
> e.g.
>
> # ./fib_nexthop_multiprefix.sh
> TEST: IPv4: host 0 to host 1, mtu 1300 [ OK ]
> TEST: IPv6: host 0 to host 1, mtu 1300 [FAIL]
>
> With -v it shows
>
> COMMAND: ip netns exec h0 /usr/sbin/ping6 -s 1350 -c5 -w5 2001:db8:101::1
> PING 2001:db8:101::1(2001:db8:101::1) 1350 data bytes
> From 2001:db8:100::64 icmp_seq=1 Packet too big: mtu=1300
>
> --- 2001:db8:101::1 ping statistics ---
> 1 packets transmitted, 0 received, +1 errors, 100% packet loss, time 0ms
>
> Route get
> 2001:db8:101::1 via 2001:db8:100::64 dev eth0 src 2001:db8:100::1 metric 1024 expires 599sec mtu 1300 pref medium
> Searching for:
> 2001:db8:101::1 from :: via 2001:db8:100::64 dev eth0 src 2001:db8:100::1 .* mtu 1300
>
> TEST: IPv6: host 0 to host 1, mtu 1300 [FAIL]
>
> So we can get the Packet too big from 2001:db8:100::64 successfully. There
> is no "from ::" anymore. I plan to fix this issue. But I can't find which
> commit changed the behavior and the client could receive Packet too big
> message with correct src address.
>
> Do you have any hints?
>
> Thanks
> Hangbin
v6.3.12:
$ sudo /mnt/hostshare/fib_nexthop_multiprefix.sh
TEST: IPv4: host 0 to host 1, mtu 1300 [ OK ]
TEST: IPv6: host 0 to host 1, mtu 1300 [ OK ]
TEST: IPv4: host 0 to host 2, mtu 1350 [ OK ]
TEST: IPv6: host 0 to host 2, mtu 1350 [ OK ]
TEST: IPv4: host 0 to host 3, mtu 1400 [ OK ]
TEST: IPv6: host 0 to host 3, mtu 1400 [ OK ]
TEST: IPv4: host 0 to host 1, mtu 1300 [ OK ]
TEST: IPv6: host 0 to host 1, mtu 1300 [ OK ]
TEST: IPv4: host 0 to host 2, mtu 1350 [ OK ]
TEST: IPv6: host 0 to host 2, mtu 1350 [ OK ]
TEST: IPv4: host 0 to host 3, mtu 1400 [ OK ]
TEST: IPv6: host 0 to host 3, mtu 1400 [ OK ]
v6.4.13 all passed as well, so it is something recent. I do not have a
6.5 or 6.6 kernels compiled at the moment.
^ permalink raw reply [flat|nested] 6+ messages in thread
* Re: selftest fib_nexthop_multiprefix failed due to route mismatch
2023-11-21 17:40 ` David Ahern
@ 2023-12-06 3:47 ` Hangbin Liu
2023-12-08 19:21 ` David Ahern
0 siblings, 1 reply; 6+ messages in thread
From: Hangbin Liu @ 2023-12-06 3:47 UTC (permalink / raw)
To: David Ahern; +Cc: netdev
On Tue, Nov 21, 2023 at 09:40:02AM -0800, David Ahern wrote:
> On 11/20/23 10:40 PM, Hangbin Liu wrote:
> > Hi David,
> >
> > Recently when run fib_nexthop_multiprefix test I saw all IPv6 test failed.
> > e.g.
> >
> > # ./fib_nexthop_multiprefix.sh
> > TEST: IPv4: host 0 to host 1, mtu 1300 [ OK ]
> > TEST: IPv6: host 0 to host 1, mtu 1300 [FAIL]
> >
> > With -v it shows
> >
> > COMMAND: ip netns exec h0 /usr/sbin/ping6 -s 1350 -c5 -w5 2001:db8:101::1
> > PING 2001:db8:101::1(2001:db8:101::1) 1350 data bytes
> > From 2001:db8:100::64 icmp_seq=1 Packet too big: mtu=1300
> >
> > --- 2001:db8:101::1 ping statistics ---
> > 1 packets transmitted, 0 received, +1 errors, 100% packet loss, time 0ms
> >
> > Route get
> > 2001:db8:101::1 via 2001:db8:100::64 dev eth0 src 2001:db8:100::1 metric 1024 expires 599sec mtu 1300 pref medium
> > Searching for:
> > 2001:db8:101::1 from :: via 2001:db8:100::64 dev eth0 src 2001:db8:100::1 .* mtu 1300
> >
> > TEST: IPv6: host 0 to host 1, mtu 1300 [FAIL]
> >
> > So we can get the Packet too big from 2001:db8:100::64 successfully. There
> > is no "from ::" anymore. I plan to fix this issue. But I can't find which
> > commit changed the behavior and the client could receive Packet too big
> > message with correct src address.
> >
> > Do you have any hints?
> >
> > Thanks
> > Hangbin
>
> v6.3.12:
>
> $ sudo /mnt/hostshare/fib_nexthop_multiprefix.sh
> TEST: IPv4: host 0 to host 1, mtu 1300 [ OK ]
> TEST: IPv6: host 0 to host 1, mtu 1300 [ OK ]
>
> v6.4.13 all passed as well, so it is something recent. I do not have a
> 6.5 or 6.6 kernels compiled at the moment.
Hi David,
I re-test this on 6.4.0 and it also failed. So this looks like an env issue
on your side?
# uname -r
6.4.0
# ./fib_nexthop_multiprefix.sh
TEST: IPv4: host 0 to host 1, mtu 1300 [ OK ]
TEST: IPv6: host 0 to host 1, mtu 1300 [FAIL]
And from the test result, it looks we should receive the Packet too big message
from r1. So look the current checking is incorrect and the "from ::" checking
should be removed.
Please fix me if I missed anything?
Thanks
Hangbin
^ permalink raw reply [flat|nested] 6+ messages in thread
* Re: selftest fib_nexthop_multiprefix failed due to route mismatch
2023-12-06 3:47 ` Hangbin Liu
@ 2023-12-08 19:21 ` David Ahern
2023-12-11 9:22 ` Hangbin Liu
0 siblings, 1 reply; 6+ messages in thread
From: David Ahern @ 2023-12-08 19:21 UTC (permalink / raw)
To: Hangbin Liu; +Cc: netdev
On 12/5/23 8:47 PM, Hangbin Liu wrote:
> On Tue, Nov 21, 2023 at 09:40:02AM -0800, David Ahern wrote:
>> On 11/20/23 10:40 PM, Hangbin Liu wrote:
>>> Hi David,
>>>
>>> Recently when run fib_nexthop_multiprefix test I saw all IPv6 test failed.
>>> e.g.
>>>
>>> # ./fib_nexthop_multiprefix.sh
>>> TEST: IPv4: host 0 to host 1, mtu 1300 [ OK ]
>>> TEST: IPv6: host 0 to host 1, mtu 1300 [FAIL]
>>>
>>> With -v it shows
>>>
>>> COMMAND: ip netns exec h0 /usr/sbin/ping6 -s 1350 -c5 -w5 2001:db8:101::1
>>> PING 2001:db8:101::1(2001:db8:101::1) 1350 data bytes
>>> From 2001:db8:100::64 icmp_seq=1 Packet too big: mtu=1300
>>>
>>> --- 2001:db8:101::1 ping statistics ---
>>> 1 packets transmitted, 0 received, +1 errors, 100% packet loss, time 0ms
>>>
>>> Route get
>>> 2001:db8:101::1 via 2001:db8:100::64 dev eth0 src 2001:db8:100::1 metric 1024 expires 599sec mtu 1300 pref medium
>>> Searching for:
>>> 2001:db8:101::1 from :: via 2001:db8:100::64 dev eth0 src 2001:db8:100::1 .* mtu 1300
>>>
>>> TEST: IPv6: host 0 to host 1, mtu 1300 [FAIL]
>>>
>>> So we can get the Packet too big from 2001:db8:100::64 successfully. There
>>> is no "from ::" anymore. I plan to fix this issue. But I can't find which
>>> commit changed the behavior and the client could receive Packet too big
>>> message with correct src address.
>>>
>>> Do you have any hints?
>>>
>>> Thanks
>>> Hangbin
>>
>> v6.3.12:
>>
>> $ sudo /mnt/hostshare/fib_nexthop_multiprefix.sh
>> TEST: IPv4: host 0 to host 1, mtu 1300 [ OK ]
>> TEST: IPv6: host 0 to host 1, mtu 1300 [ OK ]
>>
>> v6.4.13 all passed as well, so it is something recent. I do not have a
>> 6.5 or 6.6 kernels compiled at the moment.
>
> Hi David,
>
> I re-test this on 6.4.0 and it also failed. So this looks like an env issue
> on your side?
>
> # uname -r
> 6.4.0
> # ./fib_nexthop_multiprefix.sh
> TEST: IPv4: host 0 to host 1, mtu 1300 [ OK ]
> TEST: IPv6: host 0 to host 1, mtu 1300 [FAIL]
>
> And from the test result, it looks we should receive the Packet too big message
> from r1. So look the current checking is incorrect and the "from ::" checking
> should be removed.
>
> Please fix me if I missed anything?
>
I ran it in a ubuntu 20.04 VM. Do not recall any specific sysctl
settings to the VM, but maybe something is different between U20.04 and
your OS
^ permalink raw reply [flat|nested] 6+ messages in thread
* Re: selftest fib_nexthop_multiprefix failed due to route mismatch
2023-12-08 19:21 ` David Ahern
@ 2023-12-11 9:22 ` Hangbin Liu
2023-12-11 13:45 ` David Ahern
0 siblings, 1 reply; 6+ messages in thread
From: Hangbin Liu @ 2023-12-11 9:22 UTC (permalink / raw)
To: David Ahern; +Cc: netdev
On Fri, Dec 08, 2023 at 12:21:55PM -0700, David Ahern wrote:
> > Hi David,
> >
> > I re-test this on 6.4.0 and it also failed. So this looks like an env issue
> > on your side?
> >
> > # uname -r
> > 6.4.0
> > # ./fib_nexthop_multiprefix.sh
> > TEST: IPv4: host 0 to host 1, mtu 1300 [ OK ]
> > TEST: IPv6: host 0 to host 1, mtu 1300 [FAIL]
> >
> > And from the test result, it looks we should receive the Packet too big message
> > from r1. So look the current checking is incorrect and the "from ::" checking
> > should be removed.
> >
> > Please fix me if I missed anything?
> >
>
> I ran it in a ubuntu 20.04 VM. Do not recall any specific sysctl
> settings to the VM, but maybe something is different between U20.04 and
> your OS
OK, I got the reason. The "from ::1" is put in rt6_fill_node()
when CONFIG_IPV6_SUBTREES is enabled. I will fix this grep issue
in my next selftests namespace conversion patch set.
Thanks
Hangbin
^ permalink raw reply [flat|nested] 6+ messages in thread
* Re: selftest fib_nexthop_multiprefix failed due to route mismatch
2023-12-11 9:22 ` Hangbin Liu
@ 2023-12-11 13:45 ` David Ahern
0 siblings, 0 replies; 6+ messages in thread
From: David Ahern @ 2023-12-11 13:45 UTC (permalink / raw)
To: Hangbin Liu; +Cc: netdev
On 12/11/23 2:22 AM, Hangbin Liu wrote:
> OK, I got the reason. The "from ::1" is put in rt6_fill_node() when
> CONFIG_IPV6_SUBTREES is enabled. I will fix this grep issue in my next
> selftests namespace conversion patch set. Thanks Hangbin
thank you for the persistence in tracking this down.
^ permalink raw reply [flat|nested] 6+ messages in thread
end of thread, other threads:[~2023-12-11 13:45 UTC | newest]
Thread overview: 6+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2023-11-21 6:40 selftest fib_nexthop_multiprefix failed due to route mismatch Hangbin Liu
2023-11-21 17:40 ` David Ahern
2023-12-06 3:47 ` Hangbin Liu
2023-12-08 19:21 ` David Ahern
2023-12-11 9:22 ` Hangbin Liu
2023-12-11 13:45 ` David Ahern
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).