freebsd-net Digest, Vol 365, Issue 4
freebsd-n...@freebsd.org
freeb...@freebsd.org
To subscribe or unsubscribe via the World Wide Web, visit
http://lists.freebsd.org/mailman/listinfo/freebsd-net
or, via email, send a message with subject or body 'help' to
freebsd-n...@freebsd.org
You can reach the person managing the list at
freebsd-...@freebsd.org
When replying, please edit your Subject line so it is more specific
than "Re: Contents of freebsd-net digest..."
Today's Topics:
1. RE: Workaround for mpd5 and 8.0 broken proxy arp? (J. English)
2. net80211 ratectl proof of concept (Rui Paulo)
3. Re: net80211 ratectl proof of concept (Rui Paulo)
4. Re: net80211 ratectl proof of concept (Bernhard Schmidt)
5. Re: net80211 ratectl proof of concept (Rui Paulo)
6. Re: net80211 ratectl proof of concept (Bernhard Schmidt)
7. Re: net80211 ratectl proof of concept (Rui Paulo)
8. Re: kern/143788: [iwi] wpa_supplicant(8) can't set privacy on
iwi interface (bsch...@FreeBSD.org)
----------------------------------------------------------------------
Message: 1
Date: Wed, 31 Mar 2010 08:57:33 -0400
From: "J. English" <j_en...@kawasaki-tn.com>
Subject: RE: Workaround for mpd5 and 8.0 broken proxy arp?
To: "'Li, Qing'" <qin...@bluecoat.com>, "'Ruslan Ermilov'"
<r...@freebsd.org>, "'Mike Tancsa'" <mi...@sentex.net>
Cc: freeb...@freebsd.org
Message-ID: <000901cad0d1$bc08b530$341a1f90$@kawasaki-tn.com>
Content-Type: text/plain; charset="us-ascii"
Li,
I am using VLANs, but not on my FreeBSD box. Mpd log does not show any
"cannot intuit interface index" errors.
Would be happy to apply the patches. I followed the links Mike provided,
but I'm unsure of what to do at that point (I did read README and UPDATING,
but I guess I'm still too new to "get it"). Also, could you provide some
guidance on how to apply the .diff patch you provided?
Not a FreeBSD expert (yet). Thanks,
Jonathon
--
-----Original Message-----
From: Li, Qing [mailto:qin...@bluecoat.com]
Sent: Tuesday, March 30, 2010 9:26 PM
To: Ruslan Ermilov; Mike Tancsa
Cc: J. English; freeb...@freebsd.org
Subject: RE: Workaround for mpd5 and 8.0 broken proxy arp?
>
> Yes, it's still broken (it's a regression compared to 7.x). A
> workaround I've found working after analyzing the newer kernel code is
> to ALWAYS use the same IP address for the local end of the tunnel as
> of the corresponding ARP capable interface.
>
Are you using VLANs?
Were you getting the following error message?
"cannot intuit interface index and type for x.x.x.x"
Could you please revert back to your original configuration
and give the following patch try (in addition to the patches
mentioned by Mike) ?
http://people.freebsd.org/~qingli/mpd-patch.diff
Thanks,
-- Qing
------------------------------
Message: 2
Date: Wed, 31 Mar 2010 14:05:52 +0100
From: Rui Paulo <rpa...@gmail.com>
Subject: net80211 ratectl proof of concept
To: Sam Leffler <s...@errno.com>, Andrew Thompson
<tho...@freebsd.org>, Weongyo Jeong <weo...@freebsd.org>, Bernhard
Schmidt <bsch...@techwires.net>
Cc: freebsd-net <freeb...@freebsd.org>
Message-ID: <D1EDB040-BB1F-41E2...@gmail.com>
Content-Type: text/plain; charset=us-ascii
Hi,
I've started developing a ratectl framework for net80211, loosely based on what DragonFly has. Right now only one driver has been ported, but I would like your feedback before continuing.
The objective is to, eventually, have all the ratectl stuff (amrr, sample, onoe(?) and rssadapt) in net80211 so all drivers can use it. We can also select which ratectl modules are built in the kernel config file.
The framework support changing the current ratectl is out of scope for this patch.
You can find the patch here:
* http://people.freebsd.org/~rpaulo/ratectl.diff
Only the ral driver and the AMRR rate control algorithms were ported.
Some comments:
o The rate control calls now dereferences several pointers and some inline functions are now real functions. I wonder how much this impacts performance and what we can do to solve it.
o I wished there was a better way to do the IEEE80211_AMRR_SUCCESS / IEEE80211_AMRR_FAILURe call.
o Some other stuff can also be `const'
o I create ieee80211_ratect.[ch] to avoid polluting other files
o I moved the AMRR parameters inside amrr_init() on purpose. The drivers we have now only specify a different interval and I plan to add export amrr_set_interval() via the ratectl framework later.
I would like very much to see this in, unless there's a strong impending argument.
Thanks,
--
Rui Paulo
------------------------------
Message: 3
Date: Wed, 31 Mar 2010 14:31:30 +0100
From: Rui Paulo <rpa...@freebsd.org>
Subject: Re: net80211 ratectl proof of concept
To: Antoine Brodin <ant...@FreeBSD.org>
Cc: freebsd-net <freeb...@freebsd.org>
Message-ID: <63B143BF-20C9-425F...@freebsd.org>
Content-Type: text/plain; charset=us-ascii
On 31 Mar 2010, at 14:24, Antoine Brodin wrote:
> On Wed, Mar 31, 2010 at 3:05 PM, Rui Paulo <rpa...@gmail.com> wrote:
>> Hi,
>> I've started developing a ratectl framework for net80211, loosely based on what DragonFly has. Right now only one driver has been ported, but I would like your feedback before continuing.
>>
>> The objective is to, eventually, have all the ratectl stuff (amrr, sample, onoe(?) and rssadapt) in net80211 so all drivers can use it. We can also select which ratectl modules are built in the kernel config file.
>> The framework support changing the current ratectl is out of scope for this patch.
>>
>> You can find the patch here:
>> * http://people.freebsd.org/~rpaulo/ratectl.diff
>>
>> Only the ral driver and the AMRR rate control algorithms were ported.
>>
>> Some comments:
>> o The rate control calls now dereferences several pointers and some inline functions are now real functions. I wonder how much this impacts performance and what we can do to solve it.
>>
>> o I wished there was a better way to do the IEEE80211_AMRR_SUCCESS / IEEE80211_AMRR_FAILURe call.
>>
>> o Some other stuff can also be `const'
>>
>> o I create ieee80211_ratect.[ch] to avoid polluting other files
>>
>> o I moved the AMRR parameters inside amrr_init() on purpose. The drivers we have now only specify a different interval and I plan to add export amrr_set_interval() via the ratectl framework later.
>>
>>
>> I would like very much to see this in, unless there's a strong impending argument.
>
> Hello,
>
> This looks great!
> Is there specific reasons to use pointers and not ints for some
> arguments of foo_tx_complete() and foo_tx_update()?
Not really. I'll probably switch them to ints at some point.
--
Rui Paulo
------------------------------
Message: 4
Date: Wed, 31 Mar 2010 16:04:15 +0200
From: Bernhard Schmidt <bsch...@techwires.net>
Subject: Re: net80211 ratectl proof of concept
To: Rui Paulo <rpa...@freebsd.org>
Cc: freebsd-net <freeb...@freebsd.org>, Antoine Brodin
<ant...@FreeBSD.org>
Message-ID: <20100331140...@mx.techwires.net>
Content-Type: text/plain; charset=us-ascii
On Wed, Mar 31, 2010 at 02:31:30PM +0100, Rui Paulo wrote:
> On 31 Mar 2010, at 14:24, Antoine Brodin wrote:
> > This looks great!
> > Is there specific reasons to use pointers and not ints for some
> > arguments of foo_tx_complete() and foo_tx_update()?
>
> Not really. I'll probably switch them to ints at some point.
Hmm, I somehow like the idea of not passing integers here. Looking at
ath_rate/sample it passes a struct. I don't know whether this can be
rewritten though.
--
Bernhard
------------------------------
Message: 5
Date: Wed, 31 Mar 2010 15:05:27 +0100
From: Rui Paulo <rpa...@freebsd.org>
Subject: Re: net80211 ratectl proof of concept
To: Bernhard Schmidt <bsch...@techwires.net>
Cc: freebsd-net <freeb...@freebsd.org>, Antoine Brodin
<ant...@FreeBSD.org>
Message-ID: <47F572EE-B5E5-4961...@freebsd.org>
Content-Type: text/plain; charset=us-ascii
On 31 Mar 2010, at 15:04, Bernhard Schmidt wrote:
> On Wed, Mar 31, 2010 at 02:31:30PM +0100, Rui Paulo wrote:
>> On 31 Mar 2010, at 14:24, Antoine Brodin wrote:
>>> This looks great!
>>> Is there specific reasons to use pointers and not ints for some
>>> arguments of foo_tx_complete() and foo_tx_update()?
>>
>> Not really. I'll probably switch them to ints at some point.
>
> Hmm, I somehow like the idea of not passing integers here. Looking at
> ath_rate/sample it passes a struct. I don't know whether this can be
> rewritten though.
It can be rewritten to pass ints, from what I understand.
--
Rui Paulo
------------------------------
Message: 6
Date: Wed, 31 Mar 2010 19:06:48 +0200
From: Bernhard Schmidt <bsch...@techwires.net>
Subject: Re: net80211 ratectl proof of concept
To: Rui Paulo <rpa...@freebsd.org>
Cc: freebsd-net <freeb...@freebsd.org>, Antoine Brodin
<ant...@FreeBSD.org>
Message-ID: <2010033117...@mx.techwires.net>
Content-Type: text/plain; charset=us-ascii
On Wed, Mar 31, 2010 at 03:05:27PM +0100, Rui Paulo wrote:
> On 31 Mar 2010, at 15:04, Bernhard Schmidt wrote:
>
> > On Wed, Mar 31, 2010 at 02:31:30PM +0100, Rui Paulo wrote:
> >> On 31 Mar 2010, at 14:24, Antoine Brodin wrote:
> >>> This looks great!
> >>> Is there specific reasons to use pointers and not ints for some
> >>> arguments of foo_tx_complete() and foo_tx_update()?
> >>
> >> Not really. I'll probably switch them to ints at some point.
> >
> > Hmm, I somehow like the idea of not passing integers here. Looking at
> > ath_rate/sample it passes a struct. I don't know whether this can be
> > rewritten though.
>
> It can be rewritten to pass ints, from what I understand.
Ok great, looks fine to me.
Do you intend to commit this in one huge chunk (quite a few drivers
affected) or create a branch for that?
--
Bernhard
------------------------------
Message: 7
Date: Wed, 31 Mar 2010 18:14:21 +0100
From: Rui Paulo <rpa...@freebsd.org>
Subject: Re: net80211 ratectl proof of concept
To: Bernhard Schmidt <bsch...@techwires.net>
Cc: freebsd-net <freeb...@freebsd.org>, Antoine Brodin
<ant...@FreeBSD.org>
Message-ID: <B3173B42-5883-45BF...@freebsd.org>
Content-Type: text/plain; charset=us-ascii
On 31 Mar 2010, at 18:06, Bernhard Schmidt wrote:
> On Wed, Mar 31, 2010 at 03:05:27PM +0100, Rui Paulo wrote:
>> On 31 Mar 2010, at 15:04, Bernhard Schmidt wrote:
>>
>>> On Wed, Mar 31, 2010 at 02:31:30PM +0100, Rui Paulo wrote:
>>>> On 31 Mar 2010, at 14:24, Antoine Brodin wrote:
>>>>> This looks great!
>>>>> Is there specific reasons to use pointers and not ints for some
>>>>> arguments of foo_tx_complete() and foo_tx_update()?
>>>>
>>>> Not really. I'll probably switch them to ints at some point.
>>>
>>> Hmm, I somehow like the idea of not passing integers here. Looking at
>>> ath_rate/sample it passes a struct. I don't know whether this can be
>>> rewritten though.
>>
>> It can be rewritten to pass ints, from what I understand.
>
> Ok great, looks fine to me.
>
> Do you intend to commit this in one huge chunk (quite a few drivers
> affected) or create a branch for that?
I wanted people to comment on the general framework.
I wasn't planning on creating a branch, because I didn't feel it was necessary. Porting AMRR code to this framework isn't hard and I can deal with any fallout.
--
Rui Paulo
------------------------------
Message: 8
Date: Wed, 31 Mar 2010 17:28:41 GMT
From: bsch...@FreeBSD.org
Subject: Re: kern/143788: [iwi] wpa_supplicant(8) can't set privacy on
iwi interface
To: K...@MATPOCKuH.Ru, bsch...@FreeBSD.org, freeb...@FreeBSD.org
Message-ID: <201003311728....@freefall.freebsd.org>
Synopsis: [iwi] wpa_supplicant(8) can't set privacy on iwi interface
State-Changed-From-To: open->closed
State-Changed-By: bschmidt
State-Changed-When: Wed Mar 31 17:27:12 UTC 2010
State-Changed-Why:
According to the submitter this issue has been fixed.
http://www.freebsd.org/cgi/query-pr.cgi?pr=143788
------------------------------
End of freebsd-net Digest, Vol 365, Issue 4
*******************************************