Reminder: NET_NEEDS_GIANT, debug.mpsafenet going away in 7.0

[Julian Elischer]

Max Laier wrote:
> [ Excess CC-list ... testers needed!!! ]
> 
> On Tuesday 17 July 2007, Robert Watson wrote:
>> Dear all:
>>
>> This is a reminder e-mail that, in the very near future, Giant
>> compatibility shims for network protocols will be removed.
> 
> <...>
> 
>> The *only* remaining case I am aware of where removing debug.mpsafenet
>> presents an issue is credential-related firewall rules (uid, gid,
>> jail).  I'm am currently in an active e-mail discussion with the
>> various firewall maintainers about how to address this issue; as the
>> implementations of these rules violate the global lock order, deadlocks
>> occur if debug.mpsafenet isn't set to 1, which causes Giant to act as a
>> guard lock preventing parallel lock acquisition in the firewall. 
>> Hopefully we will have this resolved, in some form, soon.
> 
> What we really need right now, is real understanding of the problem (if 
> there even is any).  So we would like to ask everybody who is able to - 
> to stress test user/group rules (in pf) or uid/gid/jail rules (in ipfw) 
> with debug.mpsafenet=1  It is normal that (in an WITNESS enabled kernel) 
> you get a LOR similar to 14-17 and 32 from [1].  Everything different to 
> those should be reported.
> 
> If you indeed get a deadlock, please let us know and provide as much 
> debugging information as you can.  DDB's "ps", "show locks", "show 
> alllocks" would be perfect, but detailed information how to repeat would 
> be a good start to already.
> 
> Thanks a lot!  If you are unable to provoke a deadlock, please let us know 
> as well.  Include a few setup details (ruleset, SMP, special sysctl 
> settings ...) so we can look for patterns.

I've not seen a deadlock, only LOR warnings.

> 
> [1] http://sources.zabbadoz.net/freebsd/lor.html
>