freebsd stable - May 2006 - em(4) device still 'freezes' on latset CVSup'd 6.x ...

I just upgraded one of my old 4.x server to 6.x, and the one annoying 
problem that I've experienced with 4.x persists in 6.x ...

... when you ifconfig alias a new IP onto an em(4) device, the tcp/ip 
stack (I'm guessing) hangs for a period of time, while my 6.x/bge0 servers 
don't hang at all, they just alias ...

----
Marc G. Fournier           Hub.Org Networking Services (http://www.hub.org)
Email: scrappy@hub.org           Yahoo!: yscrappy              ICQ: 7615664

Marc G. Fournier wrote:
> I just upgraded one of my old 4.x server to 6.x, and the one annoying
> problem that I've experienced with 4.x persists in 6.x ...
> 
> ... when you ifconfig alias a new IP onto an em(4) device, the tcp/ip
> stack (I'm guessing) hangs for a period of time, while my 6.x/bge0
> servers don't hang at all, they just alias ...
You know I noticed that last night, just thought I was being stupid.

    Steve


===============================================This e.mail is private and
confidential between Multiplay (UK) Ltd. and the person or entity to whom it is
addressed. In the event of misdirection, the recipient is prohibited from using,
copying, printing or otherwise disseminating it or any information contained in
it.

In the event of misdirection, illegible or incomplete transmission please
telephone (023) 8024 3137
or return the E.mail to postmaster@multiplay.co.uk.

On 5/3/06, Marc G. Fournier <scrappy@hub.org>
wrote:
>
> I just upgraded one of my old 4.x server to 6.x, and the one annoying
> problem that I've experienced with 4.x persists in 6.x ...
>
> ... when you ifconfig alias a new IP onto an em(4) device, the tcp/ip
> stack (I'm guessing) hangs for a period of time, while my 6.x/bge0
servers
> don't hang at all, they just alias ...
This bug seems to keep coming back. There have been at least 3 PRs
about it, kern/47648, kern/81147, and kern/90167 .

I thought it was fixed with the information in kern/47648, but as it
turns out, it's still present in 4.11-RELEASE (just confirmed this). I
guess the tests I performed in kern/47648 on 4.9 did not cover all
scenarios, or there was a regression between 4.9+those minor patches
and 4.11.

I don't know the protocol for this -- would the old PRs be re-opened,
or should a new PR be filed for 6.x?

Also, as an aside ... 6.x em(4) still isn't sending up the arp broadcasts, 
so IPs moved from one machine to this one aren't having their routes 
appropriately updated ... again, please note, that moving an IP from any 
machine to a *non* em(4) interface does update the routes properly and 
instantly ...

On Wed, 3 May 2006, Marc G. Fournier wrote:
>
> I just upgraded one of my old 4.x server to 6.x, and the one annoying
problem
> that I've experienced with 4.x persists in 6.x ...
>
> ... when you ifconfig alias a new IP onto an em(4) device, the tcp/ip stack
> (I'm guessing) hangs for a period of time, while my 6.x/bge0 servers
don't
> hang at all, they just alias ...
>
> ----
> Marc G. Fournier           Hub.Org Networking Services (http://www.hub.org)
> Email: scrappy@hub.org           Yahoo!: yscrappy              ICQ: 7615664
> _______________________________________________
> freebsd-stable@freebsd.org mailing list
> http://lists.freebsd.org/mailman/listinfo/freebsd-stable
> To unsubscribe, send any mail to
"freebsd-stable-unsubscribe@freebsd.org"
>
----
Marc G. Fournier           Hub.Org Networking Services (http://www.hub.org)
Email: scrappy@hub.org           Yahoo!: yscrappy              ICQ: 7615664

Technically it's not routes that are not being updated, but a stale  
(outdated) ARP cache on the other hosts. The system with the new  
alias'ed IP needs to do a gratuitous ARP (broadcast ARP for it's own  
IP). As an intermediate solution you could flush the ARP cache on the  
hosts with stale cache (usually a router or L3 switch on the subnet).  
Sounds like a bug in the em driver. Anybody do sniffing to see if it  
does send out ARPs? If not I can test on one of our Dell 2850's with  
em's.

Regards,
Rutger Bevaart
> Date: Wed, 3 May 2006 21:43:39 -0300 (ADT)
> From: "Marc G. Fournier" <scrappy@hub.org>
> Subject: Re: em(4) device still 'freezes' on latset CVSup'd 6.x
...
> To: freebsd-stable@freebsd.org
> Message-ID: <20060503214233.G1147@ganymede.hub.org>
> Content-Type: TEXT/PLAIN; charset=US-ASCII; format=flowed
>
> Also, as an aside ... 6.x em(4) still isn't sending up the arp  
> broadcasts,
> so IPs moved from one machine to this one aren't having their routes
> appropriately updated ... again, please note, that moving an IP  
> from any
> machine to a *non* em(4) interface does update the routes properly and
> instantly ...