The third Linux Wireless mini-summit will occur on Tuesday July 22,
2008 in Ottawa, just before the Ottawa Linux Symposium. OLS is
generously supporting Linux mini-summits by providing the room and
projector (http://www.linuxsymposium.org/2008/minisummits.php).
Several vendors are supporting the event by sponsoring their own
employees to attend, and Canonical and Intel have both helped with
travel funds for key community participants as well.
Although it is already a bit late, I would like to re-open discussions
on our agenda for the meeting. Based on current upstream topics and
those persons whom I expect to be attending, I have cobbled together
an agenda proposal below. Please review it and feel free to suggest
amendments or alterations, or to elaborate or ask questions about my
proposed schedule of topics.
I look forward to seeing many of you in Ottawa!
John
P.S. If you intend to attend the event in Ottawa, please send me a
private reply to this note -- even if you believe that I already know
that you are coming! :-)
---
3rd Linux Wireless Summit
Ottawa, Canada
22 July 2008
Proposed Agenda
0830 Meet & Greet
Introductions and opening remarks
0900 Status Reports
Overview of where things stand today
- 802.11n
- QoS/Multiqueue
- Mesh mode (802.11s)
- RF kill
1015 Break
1100 Vendor/Distro Round Table
Representatives from hardware vendors and Linux distributions
discuss issues around cooperation between themselves and the
development community, what we are doing well, and what problems we
are causing for each other.
1200 Lunch/Break (sorry, no funding)
1300 Current Work
Short presentations/discussions about work in progress
- encrypted management frames (802.11w)
- regulatory enforcement (802.11d, CRDA)
- virtual hardware (mac80211_hwsim)
1445 Break
1500 Future work
Short presentations/discussions about work that has previously
been suggested or requested
- cfg80211/nl80211
- time to push to replace WEXT?
- tool enhancements and/or userland support?
- power management (suspend/resume)
- lib80211 (code sharing w/ full MAC drivers)
- rate scaling algorithms
- minstrel
- others?
- userland MLME
- userland driver implementation
1645 Break
1700 Maintainence Issues
Round table discussion of issues related to maintenance of
the wireless trees
- trivial/cleanup patches
- testing or other requirements?
- should they have lower priority?
- should they require ACKs from primary author?
- submissions to -stable
- who should drive it?
- what policies?
- rate scaling algs
- driver-specific ok? what limitations?
- build policies (part of mac80211.ko?)
- questions/suggestions/complaints
1800 Dinner/Break (sorry, no funding)
2000 BoF/Break-out Sessions
Impromptu and/or unscheduled discussions as desired by group
participants
--
John W. Linville
[email protected]
One more adjustment to account for change in the lunch schedule.
John
---
3rd Linux Wireless Summit
Les Suites Hotel, Byward Suite
Ottawa, Canada
22 July 2008
Proposed Agenda
0830 Meet & Greet
Introductions and opening remarks
0900 Status Reports
Overview of where things stand today
- QoS/Multiqueue
- 802.11n
- spectrum management
- Mesh mode (802.11s)
- RF kill
- driver status reports
1000 Break
1030 Current Work
Short presentations/discussions about work in progress
- access point mode
- cfg80211/nl80211
- time to push to replace WEXT?
- tool enhancements and/or userland support?
- regulatory enforcement (802.11d, CRDA)
- encrypted management frames (802.11w)
- virtual hardware (mac80211_hwsim)
- WiMAX
1200 Vendor/Distro Round Table
Representatives from hardware vendors and Linux distributions
discuss issues around cooperation between themselves and the
development community, what we are doing well, and what problems we
are causing for each other.
1230 Lunch/Break (provided)
1400 Future work
Short presentations/discussions about work that has previously
been suggested or requested
- power management
- suspend/resume
- power savings
- lib80211 (code sharing w/ full MAC drivers)
- rate scaling algorithms
- minstrel
- others?
- userland MLME
- userland driver implementation
- organized testing (anyone have a proposal?)
1530 BoF/Break-out/BarCamp Session Proposals
1600 Break
1630 BoF/Break-out/BarCamp Sessions
Impromptu and/or unscheduled discussions as desired by group
participants
1800 BoF/Break-out/BarCamp Session Reports
1830 Maintainence Issues
Round table discussion of issues related to maintenance of
the wireless trees
- trivial/cleanup patches
- testing or other requirements?
- should they have lower priority?
- should they require ACKs from primary author?
- submissions to -stable
- who should drive it?
- what policies?
- rate scaling algs
- driver-specific ok? what limitations?
- build policies (part of mac80211.ko?)
- questions/suggestions/complaints
1900 Adjourn
--
John W. Linville
[email protected]
On Wed, Jul 16, 2008 at 11:48 AM, John W. Linville
<[email protected]> wrote:
> - regulatory enforcement (802.11d, CRDA)
I'll cover this. If we can't cover this well in the time frame
provided we can discuss this over the course of the week in impromptu
sessions.
Luis
On Tue, 2008-07-15 at 14:26 -0400, John W. Linville wrote:
> 0900 Status Reports
>
> Overview of where things stand today
>
> - 802.11n
>
> - QoS/Multiqueue
It might make sense to do these the other way around since 11n depends
on the MQ stuff. I can talk about davem's recent changes that remove
most of the code, but other than that I think we're pretty much set.
> 1030 Current Work
>
> Short presentations/discussions about work in progress
>
> - access point mode
>
> - cfg80211/nl80211
> - time to push to replace WEXT?
> - tool enhancements and/or userland support?
Both of those are pretty much stalled now, apparently because I was the
only one investing time into them and am no longer doing anything at the
moment. I can give a five-minute overview over what we need/want for
each of those I guess.
johannes
On Tue, 2008-07-15 at 13:31 -0400, Nicolas Pitre wrote:
> On Tue, 15 Jul 2008, Dan Williams wrote:
>=20
> > On Thu, 2008-07-10 at 18:15 +0200, Michael Buesch wrote:
> > > On Thursday 10 July 2008 18:12:32 Stefanik G=C3=A1bor wrote:
> > > > Some other topics which might be of interest/good to include:
> > > > - New chipsets needing drivers (SiS 16x, Marvell, etc.)
> > >=20
> > > I have no idea what to discuss about non-existing drivers, really=
=2E
> > > Something like: "Do we need that driver? Everyone votes YES and i=
s happy"?
> >=20
> > Maybe during the week we could organize some hackfests to work on t=
he
> > Marvell/SiS/etc mac80211 drivers? Everyone bring their favorite
> > non-working Marvell 88w83xx card and we make mrv8k our bitch or
> > something like that.
>=20
> FYI...
>=20
> Lennert Buytenhek and I are writing a GPL driver for mv8300 class of=20
> Marvell WIFI chips right now, present and future versions. This is=20
> Marvell work as we work for Marvell. I have access to firmware sourc=
es,=20
> etc. but obviously cannot share them.
>=20
> We both will be in Ottawa next week.
Great! Lets definitely sit down and talk then. There is already some
GPL-ed source code for 83xx chips out there but it always used net80211=
,
not mac80211. So the mrv8k effort is meant to write a mac80211 driver
for 8310/8335/etc and perhaps TopDog chips too since they don't seem
that much different from a driver perspective.
Some of the issues with the current Marvell 83xx GPL drivers that have
gotten out are (a) code style, (b) lots of commented out stuff that's
never used, (c) OS abstraction that's unnecessary on Linux, (d) doesn't
use mac80211, and (e) uses private ioctls. So hopefully you're writing
a new 83xx driver from scratch :)
We cleaned up a lot of stuff for the OLPC 8388 libertas driver, but
unfortunately we never were able to get the Marvell 8388 team onboard,
and so they just went off and kept working on their driver and didn't
really move the upstream libertas driver forward at all.
Dan
On Tue, 15 Jul 2008, Dan Williams wrote:
> Some of the issues with the current Marvell 83xx GPL drivers that have
> gotten out are (a) code style, (b) lots of commented out stuff that's
> never used, (c) OS abstraction that's unnecessary on Linux, (d) doesn't
> use mac80211, and (e) uses private ioctls. So hopefully you're writing
> a new 83xx driver from scratch :)
Yes, this is all from scratch. The objective for this work is to be
"mainlinable".
> We cleaned up a lot of stuff for the OLPC 8388 libertas driver, but
> unfortunately we never were able to get the Marvell 8388 team onboard,
> and so they just went off and kept working on their driver and didn't
> really move the upstream libertas driver forward at all.
Libertas is a rather different beast.
Nicolas
On Wed, Jul 16, 2008 at 02:48:03PM -0400, John W. Linville wrote:
> 3rd Linux Wireless Summit
> Les Suites Hotel, Byward Suite
> Ottawa, Canada
> 22 July 2008
Just to emphasize, please note the location. The mini-summit will
be held at Les Suites, _not_ at the Congress Centre.
John
--
John W. Linville
[email protected]
On Wed, Jul 16, 2008 at 11:40 AM, Johannes Berg
<[email protected]> wrote:
> On Tue, 2008-07-15 at 14:26 -0400, John W. Linville wrote:
>
>> 0900 Status Reports
>>
>> Overview of where things stand today
>>
>> - 802.11n
>>
>> - QoS/Multiqueue
>
> It might make sense to do these the other way around since 11n depends
> on the MQ stuff. I can talk about davem's recent changes that remove
> most of the code, but other than that I think we're pretty much set.
>
>> 1030 Current Work
>>
>> Short presentations/discussions about work in progress
>>
>> - access point mode
>>
>> - cfg80211/nl80211
>> - time to push to replace WEXT?
>> - tool enhancements and/or userland support?
>
> Both of those are pretty much stalled now, apparently because I was the
> only one investing time into them and am no longer doing anything at the
> moment. I can give a five-minute overview over what we need/want for
> each of those I guess.
I believe as WEXT limitations are more visible mainly in the context
of fullmac chips.
A discussion that defining new interface or it course will be very
important. cfg80211 came a bit ahead of it's time and I believe
actually now we can spin it again.
Tomas
On Mon, Jul 14, 2008 at 06:05:13PM +0300, Guy Cohen wrote:
> I think that probably the most burning topics were a little diminished
> in our agenda, and should be put higher:
> - cfg80211/nl80211 - much of the current wext begins to fall behind
> the emerging capabilities of the mac80211 and moreover - from future
> abilities and needs of future fullmac devices. The functionality and
> ABI to userspace must be finalized as soon as possible. The status of
> this is very unclear and should be treated more thoroughly, and it
> seems it can't be shared with 5 other topics in less than 2 hours.
FWIW, I listed this one first under Future Work.
> - fullmac support - seems to be very relevant in the next couple of
> years with the integration of wifi into low power devices running
> linux.
Not sure exactly what you mean with this one. Are you talking about
the lib80211 line item? Or enhancements to the mac80211 component
to support full(er) MAC designs?
> I think that these two topics should take precedence over other topics
> that although important consume more time in the agenda, and are
> currently less likely to have impact on the average users such as 11s
> and 11w.
>
> it would be good if these topics will get substantial time slot in the agenda.
There are lots of items listed, and I'm not sure we will cover all
of them. If there are any that you would like to suggest be dropped
completely then please do so.
That said, I don't think we have enough time for real working sessions
on Tuesday. I think it would be better just to touch on a variety
of topics that can benefit from some face-to-face discussions.
>From that perspective, I think some status reports and some brief
proposals are most appropriate.
I'll post a revised (mostly just augmented) agenda proposal today.
John
--
John W. Linville
[email protected]
Tomas Winkler <[email protected]> writes:
>> Although it is already a bit late, I would like to re-open discussions
>> on our agenda for the meeting. Based on current upstream topics and
>> those persons whom I expect to be attending, I have cobbled together
>> an agenda proposal below. Please review it and feel free to suggest
>> amendments or alterations, or to elaborate or ask questions about my
>> proposed schedule of topics.
>>
>> I look forward to seeing many of you in Ottawa!
>
> One more topic would be probably
> power management/power save - connection to system power/power supply
> (Battery/AC)
I'm also very interested about this. mac80211 is lacking any support
for Power Save Mode and I think we should fix that. At least having an
interface to enable and disable PSM would be a start, but we could go
even further than that.
I'm just worried if one day is enough for us. The agenda is getting
quite long already.
--
Kalle Valo
On Thursday 10 July 2008 18:12:32 Stefanik G=E1bor wrote:
> Some other topics which might be of interest/good to include:
> - New chipsets needing drivers (SiS 16x, Marvell, etc.)
I have no idea what to discuss about non-existing drivers, really.
Something like: "Do we need that driver? Everyone votes YES and is happ=
y"?
> - Radiotap enhancements
What radiotap enhancements specifically? If it's just a field or a bit
to add, just mail the radiotap people. I think there's no need to discu=
ss that
(except for with the radiotap people on their mailinglist).
> - Userspace MLME: Plans and status
Already on the list.
--=20
Greetings Michael.
On Thu, 2008-07-10 at 09:21 +0200, Holger Schurig wrote:
> Suspend/Resume?
Umm? It's on his list.
johannes
On Thu, Jul 10, 2008 at 1:22 PM, Johannes Berg
<[email protected]> wrote:
>
>> 0900 Status Reports
>> Overview of where things stand today
>> - 802.11n
>> - QoS/Multiqueue
>> - Mesh mode (802.11s)
>> - RF kill
>
> And spectrum management (Tomas) please.
>
> We've seen initial patches that add a bit of infrastructure, but aren't
> using these things yet.
Hope to submit all the patches before symposium.
The whole package is already visible in iwl5000 branch of
iwlwifi-2.6.git Need to clean it up and submit.
Thanks
Tomas
> 0900 Status Reports
> Overview of where things stand today
> - 802.11n
> - QoS/Multiqueue
> - Mesh mode (802.11s)
> - RF kill
And spectrum management (Tomas) please.
> 1300 Current Work
>
> Short presentations/discussions about work in progress
> - encrypted management frames (802.11w)
> - regulatory enforcement (802.11d, CRDA)
> - virtual hardware (mac80211_hwsim)
or here :)
We've seen initial patches that add a bit of infrastructure, but aren't
using these things yet.
johannes
> > Not sure exactly what you mean with this one. Are you talking about
> > the lib80211 line item? Or enhancements to the mac80211 component
> > to support full(er) MAC designs?
>
> I hope just lib80211. We shouldn't be extending mac80211 to fullmac or
> quasi-fullmac parts. But we can certainly share a lot of stuff.
> lib80211 is where I see any additional fullmac effort going.
Just a minor point wrt. lib80211: I see little use in creating yet
another module since hopefully soon enough fullmac chipsets will bind
cfg80211 anyway so all of that stuff might as well be in there.
johannes
Hello,
Linux wireless is, without any doubt, a very important subject.
Many people show interest in it.
It is good to hear that the third Linux Wireless mini-summit will
occur at OLS, which will give opportunity to many people (which some
of them don't know much about linux wireless) to attend it (whereas
the two last wireless mini-summits were not part of OLS, as far as I
know).
I want to add that It would be great if part of it (or hopefully, all
of it) will be videoed and put on the web. I am sure that many
(including me) will be glad to watch it. I am sure that making an
effort like this is important and is of interest for the linux
wireless community.
Regards,
Mark Ryden
On Wed, Jul 9, 2008 at 10:09 PM, John W. Linville
<[email protected]> wrote:
> The third Linux Wireless mini-summit will occur on Tuesday July 22,
> 2008 in Ottawa, just before the Ottawa Linux Symposium. OLS is
> generously supporting Linux mini-summits by providing the room and
> projector (http://www.linuxsymposium.org/2008/minisummits.php).
> Several vendors are supporting the event by sponsoring their own
> employees to attend, and Canonical and Intel have both helped with
> travel funds for key community participants as well.
>
> Although it is already a bit late, I would like to re-open discussions
> on our agenda for the meeting. Based on current upstream topics and
> those persons whom I expect to be attending, I have cobbled together
> an agenda proposal below. Please review it and feel free to suggest
> amendments or alterations, or to elaborate or ask questions about my
> proposed schedule of topics.
>
> I look forward to seeing many of you in Ottawa!
>
> John
>
> P.S. If you intend to attend the event in Ottawa, please send me a
> private reply to this note -- even if you believe that I already know
> that you are coming! :-)
>
> ---
>
> 3rd Linux Wireless Summit
> Ottawa, Canada
> 22 July 2008
>
> Proposed Agenda
>
> 0830 Meet & Greet
>
> Introductions and opening remarks
>
> 0900 Status Reports
>
> Overview of where things stand today
>
> - 802.11n
>
> - QoS/Multiqueue
>
> - Mesh mode (802.11s)
>
> - RF kill
>
> 1015 Break
>
> 1100 Vendor/Distro Round Table
>
> Representatives from hardware vendors and Linux distributions
> discuss issues around cooperation between themselves and the
> development community, what we are doing well, and what problems we
> are causing for each other.
>
> 1200 Lunch/Break (sorry, no funding)
>
> 1300 Current Work
>
> Short presentations/discussions about work in progress
>
> - encrypted management frames (802.11w)
>
> - regulatory enforcement (802.11d, CRDA)
>
> - virtual hardware (mac80211_hwsim)
>
> 1445 Break
>
> 1500 Future work
>
> Short presentations/discussions about work that has previously
> been suggested or requested
>
> - cfg80211/nl80211
> - time to push to replace WEXT?
> - tool enhancements and/or userland support?
>
> - power management (suspend/resume)
>
> - lib80211 (code sharing w/ full MAC drivers)
>
> - rate scaling algorithms
> - minstrel
> - others?
>
> - userland MLME
>
> - userland driver implementation
>
> 1645 Break
>
> 1700 Maintainence Issues
>
> Round table discussion of issues related to maintenance of
> the wireless trees
>
> - trivial/cleanup patches
> - testing or other requirements?
> - should they have lower priority?
> - should they require ACKs from primary author?
>
> - submissions to -stable
> - who should drive it?
> - what policies?
>
> - rate scaling algs
> - driver-specific ok? what limitations?
> - build policies (part of mac80211.ko?)
>
> - questions/suggestions/complaints
>
> 1800 Dinner/Break (sorry, no funding)
>
> 2000 BoF/Break-out Sessions
>
> Impromptu and/or unscheduled discussions as desired by group
> participants
>
> --
> John W. Linville
> [email protected]
> --
> To unsubscribe from this list: send the line "unsubscribe netdev" in
> the body of a message to [email protected]
> More majordomo info at http://vger.kernel.org/majordomo-info.html
>
On Wed, Jul 16, 2008 at 11:48 AM, John W. Linville
<[email protected]> wrote:
> I'm still looking for more/different topics
A couple I can think of:
* Compat-wireless, this can be discussed after new MQ patch. We may
have to freeze mac80211 for old kernels for compat-wireless after the
MQ patch goes through and only leave in it fixes moving forward. This
consideration is because porting the MQ patch may be very difficult to
maintain. It was expected though that the first MQ patch could not be
ported though so we'll see, we just might be able to port this
properly. We can talk about this further at the summit.
* Testing -- lets get serious about automation of testing. The Orbit
grid is still available, we just need to put together scripts and test
scenarios.
* Rate control cleanup -- this needs to a good review and cleanup.
Probably not high on our priority list though. We can talk about this
further during the "rate scaling algs" section on the agenda.
Luis
On Wed, Jul 9, 2008 at 10:09 PM, John W. Linville
<[email protected]> wrote:
> The third Linux Wireless mini-summit will occur on Tuesday July 22,
> 2008 in Ottawa, just before the Ottawa Linux Symposium. OLS is
> generously supporting Linux mini-summits by providing the room and
> projector (http://www.linuxsymposium.org/2008/minisummits.php).
> Several vendors are supporting the event by sponsoring their own
> employees to attend, and Canonical and Intel have both helped with
> travel funds for key community participants as well.
>
> Although it is already a bit late, I would like to re-open discussions
> on our agenda for the meeting. Based on current upstream topics and
> those persons whom I expect to be attending, I have cobbled together
> an agenda proposal below. Please review it and feel free to suggest
> amendments or alterations, or to elaborate or ask questions about my
> proposed schedule of topics.
>
> I look forward to seeing many of you in Ottawa!
One more topic would be probably
power management/power save - connection to system power/power supply
(Battery/AC)
Jouni Malinen <[email protected]> writes:
> On Thu, Jul 10, 2008 at 05:30:47PM +0300, Kalle Valo wrote:
>
>> I'm just worried if one day is enough for us. The agenda is getting
>> quite long already.
>
> There should be plenty of time to continue discussions after (or even
> before) the mini-summit since I would assume most people stay for OLS.
Yes, you are right. I guess there are vacant rooms during the OLS
which we can use, if needed.
> For example, I will be in Ottawa for the full week (July 18-27) and
> will be available for discussions related to wireless topics both
> before and after the mini-summit itself.
I'm also very interested in any discussions. I will be arriving on the
20th and leaving on the 27th of July.
--
Kalle Valo
On Thu, 2008-07-10 at 18:15 +0200, Michael Buesch wrote:
> On Thursday 10 July 2008 18:12:32 Stefanik G=C3=A1bor wrote:
> > Some other topics which might be of interest/good to include:
> > - New chipsets needing drivers (SiS 16x, Marvell, etc.)
>=20
> I have no idea what to discuss about non-existing drivers, really.
> Something like: "Do we need that driver? Everyone votes YES and is ha=
ppy"?
Maybe during the week we could organize some hackfests to work on the
Marvell/SiS/etc mac80211 drivers? Everyone bring their favorite
non-working Marvell 88w83xx card and we make mrv8k our bitch or
something like that.
Dan
> > - Radiotap enhancements
>=20
> What radiotap enhancements specifically? If it's just a field or a bi=
t
> to add, just mail the radiotap people. I think there's no need to dis=
cuss that
> (except for with the radiotap people on their mailinglist).
>=20
> > - Userspace MLME: Plans and status
>=20
> Already on the list.
>=20
I adjusted the times based on the schedule for lunch and coffee breaks
from the OLS crew. I also adjusted the length of some slots, and
move the BoF/Break-Out/BarCamp section before adjourning for dinner.
I'm still looking for more/different topics and for people to
explicitly claim the ones on the list...speak up!
John
---
3rd Linux Wireless Summit
Les Suites Hotel, Byward Suite
Ottawa, Canada
22 July 2008
Proposed Agenda
0830 Meet & Greet
Introductions and opening remarks
0900 Status Reports
Overview of where things stand today
- QoS/Multiqueue
- 802.11n
- spectrum management
- Mesh mode (802.11s)
- RF kill
- driver status reports
1000 Break
1030 Current Work
Short presentations/discussions about work in progress
- access point mode
- cfg80211/nl80211
- time to push to replace WEXT?
- tool enhancements and/or userland support?
- regulatory enforcement (802.11d, CRDA)
- encrypted management frames (802.11w)
- virtual hardware (mac80211_hwsim)
- WiMAX
1200 Vendor/Distro Round Table
Representatives from hardware vendors and Linux distributions
discuss issues around cooperation between themselves and the
development community, what we are doing well, and what problems we
are causing for each other.
1230 Lunch/Break (provided by HP)
1330 Future work
Short presentations/discussions about work that has previously
been suggested or requested
- power management
- suspend/resume
- power savings
- lib80211 (code sharing w/ full MAC drivers)
- rate scaling algorithms
- minstrel
- others?
- userland MLME
- userland driver implementation
- organized testing (anyone have a proposal?)
1500 BoF/Break-out/BarCamp Session Proposals
1530 Break
1600 BoF/Break-out/BarCamp Sessions
Impromptu and/or unscheduled discussions as desired by group
participants
1730 BoF/Break-out/BarCamp Session Reports
1800 Maintainence Issues
Round table discussion of issues related to maintenance of
the wireless trees
- trivial/cleanup patches
- testing or other requirements?
- should they have lower priority?
- should they require ACKs from primary author?
- submissions to -stable
- who should drive it?
- what policies?
- rate scaling algs
- driver-specific ok? what limitations?
- build policies (part of mac80211.ko?)
- questions/suggestions/complaints
1830 Adjourn
--
John W. Linville
[email protected]
On Thu, Jul 10, 2008 at 5:31 PM, Johannes Berg
<[email protected]> wrote:
>
>> 1500 Future work
>
> Maybe something about testing? Or I'll do a testing brainstorming
> session after dinner.
>
> Obviously, testing at orbit never worked out, but it'd be good to have
> _something_ other than everybody testing their thing locally. Ties in
> with the virtual hardware, maybe?
>
>From the driver perspective it would be loop back phy (tx/rx), The
peer side might be running on virtual HW. Currently it's not easy to
establish automated regression over the air. Good portion of the flows
can be tested this way.
I'm also looking for easier bug reporting for unexperienced users
maybe some uevent/HAL popping to application and emails it out/. Not
sure really how to handle this and not create spams. I'm just
expecting more real 'end users'.
Thanks
Tomas
>
> > The third Linux Wireless mini-summit will occur on Tuesday July 22,
> > 2008 in Ottawa, just before the Ottawa Linux Symposium. OLS is
> > generously supporting Linux mini-summits by providing the room and
> > projector (http://www.linuxsymposium.org/2008/minisummits.php).
> > Several vendors are supporting the event by sponsoring their own
> > employees to attend, and Canonical and Intel have both helped with
> > travel funds for key community participants as well.
> >
> > Although it is already a bit late, I would like to re-open discussions
> > on our agenda for the meeting. Based on current upstream topics and
> > those persons whom I expect to be attending, I have cobbled together
> > an agenda proposal below. Please review it and feel free to suggest
> > amendments or alterations, or to elaborate or ask questions about my
> > proposed schedule of topics.
> >
> > I look forward to seeing many of you in Ottawa!
>
> One more topic would be probably
> power management/power save - connection to system power/power supply
> (Battery/AC)
brief discussion about AP mode current and future plans would be
appreciated, maybe as part of the MLME topic.
On Wednesday 09 July 2008, John W. Linville wrote:
> The third Linux Wireless mini-summit will occur on Tuesday July 22,
> 2008 in Ottawa, just before the Ottawa Linux Symposium. OLS is
> generously supporting Linux mini-summits by providing the room and
> projector (http://www.linuxsymposium.org/2008/minisummits.php).
> Several vendors are supporting the event by sponsoring their own
> employees to attend, and Canonical and Intel have both helped with
> travel funds for key community participants as well.
>
> Although it is already a bit late, I would like to re-open discussions
> on our agenda for the meeting. Based on current upstream topics and
> those persons whom I expect to be attending, I have cobbled together
> an agenda proposal below. Please review it and feel free to suggest
> amendments or alterations, or to elaborate or ask questions about my
> proposed schedule of topics.
Please add a note about WiMAX. We'd like to give a quick intro to the
technology and the infrastructure work we are doing for it.
--
Inaky
> 1500 Future work
Maybe something about testing? Or I'll do a testing brainstorming
session after dinner.
Obviously, testing at orbit never worked out, but it'd be good to have
_something_ other than everybody testing their thing locally. Ties in
with the virtual hardware, maybe?
johannes
On Mon, 21 Jul 2008, Dan Williams wrote:
> On Tue, 2008-07-15 at 14:14 -0400, Nicolas Pitre wrote:
> > On Tue, 15 Jul 2008, Dan Williams wrote:
> >
> > > Some of the issues with the current Marvell 83xx GPL drivers that have
> > > gotten out are (a) code style, (b) lots of commented out stuff that's
> > > never used, (c) OS abstraction that's unnecessary on Linux, (d) doesn't
> > > use mac80211, and (e) uses private ioctls. So hopefully you're writing
> > > a new 83xx driver from scratch :)
> >
> > Yes, this is all from scratch. The objective for this work is to be
> > "mainlinable".
>
> BTW, any chance on putting up a git repo somewhere so others can poke
> around with it? I'm quite interested in hacking on it.
Eventually, of course. Let's discuss f2f tomorrow.
Nicolas
On Wed, 2008-07-16 at 10:24 +0200, Johannes Berg wrote:
> > > Not sure exactly what you mean with this one. Are you talking about
> > > the lib80211 line item? Or enhancements to the mac80211 component
> > > to support full(er) MAC designs?
> >
> > I hope just lib80211. We shouldn't be extending mac80211 to fullmac or
> > quasi-fullmac parts. But we can certainly share a lot of stuff.
> > lib80211 is where I see any additional fullmac effort going.
>
> Just a minor point wrt. lib80211: I see little use in creating yet
> another module since hopefully soon enough fullmac chipsets will bind
> cfg80211 anyway so all of that stuff might as well be in there.
worksforme, I don't really care where that stuff goes as long as we get
it in one place :)
Dan
I think that probably the most burning topics were a little diminished
in our agenda, and should be put higher:
- cfg80211/nl80211 - much of the current wext begins to fall behind
the emerging capabilities of the mac80211 and moreover - from future
abilities and needs of future fullmac devices. The functionality and
ABI to userspace must be finalized as soon as possible. The status of
this is very unclear and should be treated more thoroughly, and it
seems it can't be shared with 5 other topics in less than 2 hours.
- fullmac support - seems to be very relevant in the next couple of
years with the integration of wifi into low power devices running
linux.
I think that these two topics should take precedence over other topics
that although important consume more time in the agenda, and are
currently less likely to have impact on the average users such as 11s
and 11w.
it would be good if these topics will get substantial time slot in the agenda.
I think that probably the most burning topics were a little diminished
in our agenda, and should be put higher:
- cfg80211/nl80211 - much of the current wext begins to fall behind
the emerging capabilities of the mac80211 and moreover - from future
abilities and needs of future fullmac devices. The functionality and
ABI to userspace must be finalized as soon as possible. The status of
this is very unclear and should be treated more thoroughly, and it
seems it can't be shared with 5 other topics in less than 2 hours.
- fullmac support - seems to be very relevant in the next couple of
years with the integration of wifi into low power devices running
linux.
I think that these two topics should take precedence over other topics
that although important consume more time in the agenda, and are
currently less likely to have impact on the average users such as 11s
and 11w.
it would be good if these topics will get substantial time slot in the agenda.
"Dan Williams" <[email protected]> writes:
> Some of the issues with the current Marvell 83xx GPL drivers that have
> gotten out are (a) code style, (b) lots of commented out stuff that's
> never used, (c) OS abstraction that's unnecessary on Linux, (d) doesn't
> use mac80211, and (e) uses private ioctls.
I think you just listed exactly the problems with all vendor WLAN
drivers.
I think Intel is the only expection here. Is there anyone else?
> So hopefully you're writing a new 83xx driver from scratch :)
And it always seems that this is the only viable way to solve it :)
(dropping lkml and netdev from CC)
--
Kalle Valo
On Tue, 2008-07-15 at 14:14 -0400, Nicolas Pitre wrote:
> On Tue, 15 Jul 2008, Dan Williams wrote:
>
> > Some of the issues with the current Marvell 83xx GPL drivers that have
> > gotten out are (a) code style, (b) lots of commented out stuff that's
> > never used, (c) OS abstraction that's unnecessary on Linux, (d) doesn't
> > use mac80211, and (e) uses private ioctls. So hopefully you're writing
> > a new 83xx driver from scratch :)
>
> Yes, this is all from scratch. The objective for this work is to be
> "mainlinable".
BTW, any chance on putting up a git repo somewhere so others can poke
around with it? I'm quite interested in hacking on it.
Dan
On Tue, 2008-07-15 at 13:42 -0400, John W. Linville wrote:
> On Mon, Jul 14, 2008 at 06:05:13PM +0300, Guy Cohen wrote:
> > I think that probably the most burning topics were a little diminished
> > in our agenda, and should be put higher:
> > - cfg80211/nl80211 - much of the current wext begins to fall behind
> > the emerging capabilities of the mac80211 and moreover - from future
> > abilities and needs of future fullmac devices. The functionality and
> > ABI to userspace must be finalized as soon as possible. The status of
> > this is very unclear and should be treated more thoroughly, and it
> > seems it can't be shared with 5 other topics in less than 2 hours.
>
> FWIW, I listed this one first under Future Work.
>
> > - fullmac support - seems to be very relevant in the next couple of
> > years with the integration of wifi into low power devices running
> > linux.
>
> Not sure exactly what you mean with this one. Are you talking about
> the lib80211 line item? Or enhancements to the mac80211 component
> to support full(er) MAC designs?
I hope just lib80211. We shouldn't be extending mac80211 to fullmac or
quasi-fullmac parts. But we can certainly share a lot of stuff.
lib80211 is where I see any additional fullmac effort going.
Dan
> > I think that these two topics should take precedence over other topics
> > that although important consume more time in the agenda, and are
> > currently less likely to have impact on the average users such as 11s
> > and 11w.
> >
> > it would be good if these topics will get substantial time slot in the agenda.
>
> There are lots of items listed, and I'm not sure we will cover all
> of them. If there are any that you would like to suggest be dropped
> completely then please do so.
>
> That said, I don't think we have enough time for real working sessions
> on Tuesday. I think it would be better just to touch on a variety
> of topics that can benefit from some face-to-face discussions.
> From that perspective, I think some status reports and some brief
> proposals are most appropriate.
>
> I'll post a revised (mostly just augmented) agenda proposal today.
>
> John
On Tue, 15 Jul 2008, Dan Williams wrote:
> On Thu, 2008-07-10 at 18:15 +0200, Michael Buesch wrote:
> > On Thursday 10 July 2008 18:12:32 Stefanik Gábor wrote:
> > > Some other topics which might be of interest/good to include:
> > > - New chipsets needing drivers (SiS 16x, Marvell, etc.)
> >
> > I have no idea what to discuss about non-existing drivers, really.
> > Something like: "Do we need that driver? Everyone votes YES and is happy"?
>
> Maybe during the week we could organize some hackfests to work on the
> Marvell/SiS/etc mac80211 drivers? Everyone bring their favorite
> non-working Marvell 88w83xx card and we make mrv8k our bitch or
> something like that.
FYI...
Lennert Buytenhek and I are writing a GPL driver for mv8300 class of
Marvell WIFI chips right now, present and future versions. This is
Marvell work as we work for Marvell. I have access to firmware sources,
etc. but obviously cannot share them.
We both will be in Ottawa next week.
Nicolas
On Thu, Jul 10, 2008 at 05:30:47PM +0300, Kalle Valo wrote:
> I'm just worried if one day is enough for us. The agenda is getting
> quite long already.
There should be plenty of time to continue discussions after (or even
before) the mini-summit since I would assume most people stay for OLS.
For example, I will be in Ottawa for the full week (July 18-27) and
will be available for discussions related to wireless topics both
before and after the mini-summit itself.
--
Jouni Malinen PGP id EFC895FA
On Tue, 2008-07-15 at 14:14 -0400, Nicolas Pitre wrote:
> On Tue, 15 Jul 2008, Dan Williams wrote:
>
> > Some of the issues with the current Marvell 83xx GPL drivers that have
> > gotten out are (a) code style, (b) lots of commented out stuff that's
> > never used, (c) OS abstraction that's unnecessary on Linux, (d) doesn't
> > use mac80211, and (e) uses private ioctls. So hopefully you're writing
> > a new 83xx driver from scratch :)
>
> Yes, this is all from scratch. The objective for this work is to be
> "mainlinable".
>
> > We cleaned up a lot of stuff for the OLPC 8388 libertas driver, but
> > unfortunately we never were able to get the Marvell 8388 team onboard,
> > and so they just went off and kept working on their driver and didn't
> > really move the upstream libertas driver forward at all.
>
> Libertas is a rather different beast.
Obviously; this was just an observation about the last time we got a
wireless driver from Marvell. I expect it to work out rather
differently this time simply on the strength of your and Lennert's
track-record, but if it's really just you two guys and you don't have
buy-in from the actual hardware and software teams inside Marvell, it's
hard to see how this wouldn't just happen again.
Dan
On Wed, Jul 9, 2008 at 9:09 PM, John W. Linville <[email protected]> wrote:
> The third Linux Wireless mini-summit will occur on Tuesday July 22,
> 2008 in Ottawa, just before the Ottawa Linux Symposium. OLS is
> generously supporting Linux mini-summits by providing the room and
> projector (http://www.linuxsymposium.org/2008/minisummits.php).
> Several vendors are supporting the event by sponsoring their own
> employees to attend, and Canonical and Intel have both helped with
> travel funds for key community participants as well.
>
> Although it is already a bit late, I would like to re-open discussions
> on our agenda for the meeting. Based on current upstream topics and
> those persons whom I expect to be attending, I have cobbled together
> an agenda proposal below. Please review it and feel free to suggest
> amendments or alterations, or to elaborate or ask questions about my
> proposed schedule of topics.
>
> I look forward to seeing many of you in Ottawa!
>
> John
>
> P.S. If you intend to attend the event in Ottawa, please send me a
> private reply to this note -- even if you believe that I already know
> that you are coming! :-)
>
> ---
>
> 3rd Linux Wireless Summit
> Ottawa, Canada
> 22 July 2008
>
> Proposed Agenda
>
> 0830 Meet & Greet
>
> Introductions and opening remarks
>
> 0900 Status Reports
>
> Overview of where things stand today
>
> - 802.11n
>
> - QoS/Multiqueue
>
> - Mesh mode (802.11s)
>
> - RF kill
>
> 1015 Break
>
> 1100 Vendor/Distro Round Table
>
> Representatives from hardware vendors and Linux distributions
> discuss issues around cooperation between themselves and the
> development community, what we are doing well, and what problems we
> are causing for each other.
>
> 1200 Lunch/Break (sorry, no funding)
>
> 1300 Current Work
>
> Short presentations/discussions about work in progress
>
> - encrypted management frames (802.11w)
>
> - regulatory enforcement (802.11d, CRDA)
>
> - virtual hardware (mac80211_hwsim)
>
> 1445 Break
>
> 1500 Future work
>
> Short presentations/discussions about work that has previously
> been suggested or requested
>
> - cfg80211/nl80211
> - time to push to replace WEXT?
> - tool enhancements and/or userland support?
>
> - power management (suspend/resume)
>
> - lib80211 (code sharing w/ full MAC drivers)
>
> - rate scaling algorithms
> - minstrel
> - others?
>
> - userland MLME
>
> - userland driver implementation
>
> 1645 Break
>
> 1700 Maintainence Issues
>
> Round table discussion of issues related to maintenance of
> the wireless trees
>
> - trivial/cleanup patches
> - testing or other requirements?
> - should they have lower priority?
> - should they require ACKs from primary author?
>
> - submissions to -stable
> - who should drive it?
> - what policies?
>
> - rate scaling algs
> - driver-specific ok? what limitations?
> - build policies (part of mac80211.ko?)
>
> - questions/suggestions/complaints
>
> 1800 Dinner/Break (sorry, no funding)
>
> 2000 BoF/Break-out Sessions
>
> Impromptu and/or unscheduled discussions as desired by group
> participants
>
> --
> John W. Linville
> [email protected]
> --
> To unsubscribe from this list: send the line "unsubscribe linux-wireless" in
> the body of a message to [email protected]
> More majordomo info at http://vger.kernel.org/majordomo-info.html
>
Some other topics which might be of interest/good to include:
- New chipsets needing drivers (SiS 16x, Marvell, etc.)
- Radiotap enhancements
- Userspace MLME: Plans and status
--
Vista: [V]iruses, [I]ntruders, [S]pyware, [T]rojans and [A]dware. :-)
I revised the agenda based on the feedback from last week.
Most changes are fairly minor, although I moved the vendor/distro
round table after lunch to give us a chance to cover some in-progress
work beforehand.
As I mentioned elsewhere in the thread, I don't think we have enough
time for work sessions on Tuesday. As others have pointed-out, there
is plenty of time for extended discussions during the days at OLS.
So I think we should keep Tuesday for more high-level design and
development process discussions.
Feel free to volunteer to cover any of the listed topics. Further
suggestions to alter or augment the agenda are, of course, still
welcome as well.
Thanks!
John
---
3rd Linux Wireless Summit
Ottawa, Canada
22 July 2008
Proposed Agenda
0830 Meet & Greet
Introductions and opening remarks
0900 Status Reports
Overview of where things stand today
- 802.11n
- QoS/Multiqueue
- spectrum management
- Mesh mode (802.11s)
- RF kill
1015 Break
1030 Current Work
Short presentations/discussions about work in progress
- access point mode
- cfg80211/nl80211
- time to push to replace WEXT?
- tool enhancements and/or userland support?
- regulatory enforcement (802.11d, CRDA)
- encrypted management frames (802.11w)
- virtual hardware (mac80211_hwsim)
- WiMAX
1200 Lunch/Break (provided by HP)
1300 Vendor/Distro Round Table
Representatives from hardware vendors and Linux distributions
discuss issues around cooperation between themselves and the
development community, what we are doing well, and what problems we
are causing for each other.
1445 Break
1500 Future work
Short presentations/discussions about work that has previously
been suggested or requested
- power management
- suspend/resume
- power savings
- lib80211 (code sharing w/ full MAC drivers)
- rate scaling algorithms
- minstrel
- others?
- userland MLME
- userland driver implementation
- organized testing?
1645 Break
1700 Maintainence Issues
Round table discussion of issues related to maintenance of
the wireless trees
- trivial/cleanup patches
- testing or other requirements?
- should they have lower priority?
- should they require ACKs from primary author?
- submissions to -stable
- who should drive it?
- what policies?
- rate scaling algs
- driver-specific ok? what limitations?
- build policies (part of mac80211.ko?)
- questions/suggestions/complaints
1800 Dinner/Break (sorry, no funding)
2000 BoF/Break-out Sessions
Impromptu and/or unscheduled discussions as desired by group
participants
--
John W. Linville
[email protected]
On Tue, 2008-07-15 at 13:58 -0400, Dan Williams wrote:
> On Tue, 2008-07-15 at 13:31 -0400, Nicolas Pitre wrote:
> > On Tue, 15 Jul 2008, Dan Williams wrote:
> >=20
> > > On Thu, 2008-07-10 at 18:15 +0200, Michael Buesch wrote:
> > > > On Thursday 10 July 2008 18:12:32 Stefanik G=C3=A1bor wrote:
> > > > > Some other topics which might be of interest/good to include:
> > > > > - New chipsets needing drivers (SiS 16x, Marvell, etc.)
> > > >=20
> > > > I have no idea what to discuss about non-existing drivers, real=
ly.
> > > > Something like: "Do we need that driver? Everyone votes YES and=
is happy"?
> > >=20
> > > Maybe during the week we could organize some hackfests to work on=
the
> > > Marvell/SiS/etc mac80211 drivers? Everyone bring their favorite
> > > non-working Marvell 88w83xx card and we make mrv8k our bitch or
> > > something like that.
> >=20
> > FYI...
> >=20
> > Lennert Buytenhek and I are writing a GPL driver for mv8300 class o=
f=20
> > Marvell WIFI chips right now, present and future versions. This is=20
> > Marvell work as we work for Marvell. I have access to firmware sou=
rces,=20
> > etc. but obviously cannot share them.
> >=20
> > We both will be in Ottawa next week.
>=20
> Great! Lets definitely sit down and talk then. There is already som=
e
> GPL-ed source code for 83xx chips out there but it always used net802=
11,
> not mac80211. So the mrv8k effort is meant to write a mac80211 drive=
r
> for 8310/8335/etc and perhaps TopDog chips too since they don't seem
> that much different from a driver perspective.
>=20
> Some of the issues with the current Marvell 83xx GPL drivers that hav=
e
> gotten out are (a) code style, (b) lots of commented out stuff that's
> never used, (c) OS abstraction that's unnecessary on Linux, (d) doesn=
't
> use mac80211, and (e) uses private ioctls. So hopefully you're writi=
ng
> a new 83xx driver from scratch :)
>=20
> We cleaned up a lot of stuff for the OLPC 8388 libertas driver, but
> unfortunately we never were able to get the Marvell 8388 team onboard=
,
> and so they just went off and kept working on their driver and didn't
> really move the upstream libertas driver forward at all.
=46orgot to finish this last bit...
So hopefully that won't happen here, since this would be pretty much th=
e
same situation up to this point. Ideally you can get the internal
Marvell developers onboard so we're all working on essentially the same
bits, exactly like Intel has done. Yeah, they develop on private trees
internally but at the end of the day, their base code is the same
upstream base code everyone else uses.
Dan