On Jun 1, 2013, at 11:35 AM, Tom Eastep <[email protected]> wrote:

> 
> On Jun 1, 2013, at 11:13 AM, Dash Four <[email protected]> wrote:
> 
>> 
>> Tom Eastep wrote:
>>> On Jun 1, 2013, at 10:39 AM, Dash Four <[email protected]> wrote:
>>> 
>>> 
>>>> I think I finally got the bastard!
>>>> 
>>>> Now, if I have the above statement in rules and have *no* other 
>>>> statements present, I am *not* getting these warnings. However, if I add 
>>>> the following:
>>>> 
>>>> rules
>>>> ~~~~~
>>>> SECTION RELATED
>>>> IFLOG(-,log1,-,accept,ACCEPT) $FW local
>>>> IFLOG(-,log1,-,accept,ACCEPT) local $FW
>>>> 
>>>> IFLOG(-,log1,-,drop,DROP) all all
>>>> 
>>>> Then I get the warnings - all 4 of them, directing me at the last 
>>>> statement line ("all all"). Now, if I comment out either of the "$FW 
>>>> local" or "local $FW" statements, then I get only 2 warnings instead. If 
>>>> I comment out the last statement, then I don't get any warnings at all.
>>>> 
>>>> So, what I think is happening is this:
>>>> 
>>>> 1. The 3 statements above do something in combination that shorewall 
>>>> doesn't like very much and issues these warnings.
>>>> 2. shorewall is telling me porkies about the erroneous line in my 
>>>> "rules" statement file (that the problem is with my last statement), 
>>>> confusing the hell out of me.
>>>> 
>>>> Over to you Tom…
>>>> 
>>> 
>>> Okay -- apply this patch for now.
>>> 
>> Before I do that, I discovered another - nastier bug:
>> 
>> actions
>> ~~~~~~~
>> IFLOG inline
>> SFLOG
>> 
>> action.SFLOG
>> ~~~~~~~~~~~~
>> ?SET p6 $6 ? $6 : @{chain}
>> ?IF $5 eq 'Drop'
>> $5
>> ?ENDIF
>> IFLOG($1,$2,$3,$4,$5) ; switch:${p6}_${7}
>> ?IF $5 && (! ($5 eq 'Drop'))
>> $5
>> ?ENDIF
>> 
>> rules
>> ~~~~~
>> SFLOG(-,-,-,-,ACCEPT,-,log_test_related=0) $FW local:+test
>> 
>> produces:
>> 
>> -A +fw2local -m set --match-set test dst -j ACCEPT
>> 
>> Please note that this is a straight ACCEPT jump with no conditional 
>> switch. The rule produced should have been:
>> 
>> -A +fw2local --condition fw2local_log_test_related -m set --match-set 
>> test dst -j ACCEPT
> 
> I'm not even sure what the semantics of applying raw input to an inline 
> invocation should be. Apply it to every entry in the action body? What if an 
> entry in the body has raw input supplied?

Never mind -- I misread the example.

-Tom

Tom Eastep        \ When I die, I want to go like my Grandfather who
Shoreline,         \ died peacefully in his sleep. Not screaming like
Washington, USA     \ all of the passengers in his car
http://shorewall.net \________________________________________________




------------------------------------------------------------------------------
Get 100% visibility into Java/.NET code with AppDynamics Lite
It's a free troubleshooting tool designed for production
Get down to code-level detail for bottlenecks, with <2% overhead.
Download for free and get started troubleshooting in minutes.
http://p.sf.net/sfu/appdyn_d2d_ap2
_______________________________________________
Shorewall-devel mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/shorewall-devel

Reply via email to