ip6tables functionality test is wanted

ip6tables functionality test is wanted

[Konstantin Khorenko]

Dear All,

working on iptables IPv6 32bit compat kernel code implementation we need to somehow test it.
i've spent some time searching for a good ip6tables test, but have to admit that did not find any.

On the other hand i guess during the development new version of "iptables" could be tested on some regression test suites.

i'd very appreciate if someone can point me some test for ip6tables we can use.
(Checking the kernel part is not important in our case - we can simply compare the resulted chains/rules after the test execution with the list taken on a native 32bit node,
so the most important thing is a list of valid ip6tables commands.)

Thank you in advance.

--
Best regards,
Konstantin Khorenko

Re: ip6tables functionality test is wanted

[Pablo Neira Ayuso]

On 06/06/11 21:33, Konstantin Khorenko wrote:
> Dear All,
> 
> working on iptables IPv6 32bit compat kernel code implementation we need
> to somehow test it.
> i've spent some time searching for a good ip6tables test, but have to
> admit that did not find any.
> 
> On the other hand i guess during the development new version of
> "iptables" could be tested on some regression test suites.

These regression tests are missing, we should have some indeed.

Rusty already considered this in nfsim, but that project was not any
longer supported (getting part of the kernel compile to compile in
userspace with every version is a daunting task).

> i'd very appreciate if someone can point me some test for ip6tables we
> can use.
> (Checking the kernel part is not important in our case - we can simply
> compare the resulted chains/rules after the test execution with the list
> taken on a native 32bit node,
> so the most important thing is a list of valid ip6tables commands.)

Sorry, we don't have them at the moment.

Re: ip6tables functionality test is wanted

[Jan Engelhardt]

On Tuesday 2011-06-07 01:36, Pablo Neira Ayuso wrote:

>On 06/06/11 21:33, Konstantin Khorenko wrote:
>> Dear All,
>> 
>> working on iptables IPv6 32bit compat kernel code implementation we need
>> to somehow test it.
>> i've spent some time searching for a good ip6tables test, but have to
>> admit that did not find any.
>> 
>> On the other hand i guess during the development new version of
>> "iptables" could be tested on some regression test suites.
>
>These regression tests are missing, we should have some indeed.
>
>Rusty already considered this in nfsim, but that project was not any
>longer supported (getting part of the kernel compile to compile in
>userspace with every version is a daunting task).
>
>> i'd very appreciate if someone can point me some test for ip6tables we
>> can use.
>> (Checking the kernel part is not important in our case - we can simply
>> compare the resulted chains/rules after the test execution with the list
>> taken on a native 32bit node,
>> so the most important thing is a list of valid ip6tables commands.)
>
>Sorry, we don't have them at the moment.

I have rule test files that were helpful during the Guided Option Parser 
implementation. I shall submit.

About compat, it is as simple as running an ELF32 executable of iptables 
to trigger its use.