Hi !
Just a little question in the light of the discussion we had at Kernel
Summit about merging drivers upstream (and here, I strongly agree with
Linus, hence my message).
I just got that new T61 laptop which happens to have an iwl4xxx chip.
The distro I installed on it (ubuntu) has a driver for it. I suspect
others do too and most users get it from some random external tree and
use it.
Thus my question, why are we about to release 2.6.23 without it ?
It doesn't seem to pull any depedency nor affect any other external
piece of code unless I'm missing something, so it's a perfect example of
what we've been discussing back then: there is just no point not merging
it at any time right ? :-)
Cheers,
Ben.
On Fri, Sep 28, 2007 at 11:39:27AM +1000, Benjamin Herrenschmidt wrote:
>
> Just a little question in the light of the discussion we had at Kernel
> Summit about merging drivers upstream (and here, I strongly agree with
> Linus, hence my message).
>
> I just got that new T61 laptop which happens to have an iwl4xxx chip.
> The distro I installed on it (ubuntu) has a driver for it. I suspect
> others do too and most users get it from some random external tree and
> use it.
>
> Thus my question, why are we about to release 2.6.23 without it ?
>
> It doesn't seem to pull any depedency nor affect any other external
> piece of code unless I'm missing something, so it's a perfect example of
> what we've been discussing back then: there is just no point not merging
> it at any time right ? :-)
Well, pulling in iwlwifi would require also pulling in the mac80211
subsystem, so it's not quite that simple (although I'm not sure what's
holding back that going into the kernel.)
I had no problem building my personal production kernel by taking
2.6.23-rc8, and doing a git pull from the everything branch in John
Linville's wireless-dev git tree. It's probably too late to pull it
for 2.6.23-rc8 (although if Linux wanted to do it it's only one git
pull command away :-), but it would be really nice if it could get
merged in for 2.6.24.
- Ted
On Fri, Sep 28, 2007 at 11:39:27AM +1000, Benjamin Herrenschmidt wrote:
> Just a little question in the light of the discussion we had at Kernel
> Summit about merging drivers upstream (and here, I strongly agree with
> Linus, hence my message).
You must not have been watching me SPAM netdev for the past two
weeks. :-)
> I just got that new T61 laptop which happens to have an iwl4xxx chip.
> The distro I installed on it (ubuntu) has a driver for it. I suspect
> others do too and most users get it from some random external tree and
> use it.
>
> Thus my question, why are we about to release 2.6.23 without it ?
>
> It doesn't seem to pull any depedency nor affect any other external
> piece of code unless I'm missing something, so it's a perfect example of
> what we've been discussing back then: there is just no point not merging
> it at any time right ? :-)
It is queued for 2.6.24. I'm not even sure it was originally posted
in time for the 2.6.23 merge window, but even if it was there was
a lot of opposition to merging it until fairly recently. In fact,
I'm sure there are still some wireless developers that are less than
happy about merging it now.
Anyway, coming soon to a kernel near you...
John
--
John W. Linville
[email protected]
>
> Well, pulling in iwlwifi would require also pulling in the mac80211
> subsystem, so it's not quite that simple (although I'm not sure what's
> holding back that going into the kernel.)
I though that was already in 2.6.23 ... my bad if I missed something
(there is definitely something there called net/mac80211)
> I had no problem building my personal production kernel by taking
> 2.6.23-rc8, and doing a git pull from the everything branch in John
> Linville's wireless-dev git tree. It's probably too late to pull it
> for 2.6.23-rc8 (although if Linux wanted to do it it's only one git
> pull command away :-), but it would be really nice if it could get
> merged in for 2.6.24.
Yes, I agree -rc8 seems to be a tad too late, I'm just surprised we
didn't get it in earlier though since it seems it's been around and
useable for some time.
Cheers,
Ben.
On Thu, 2007-09-27 at 22:30 -0400, John W. Linville wrote:
> > It doesn't seem to pull any depedency nor affect any other external
> > piece of code unless I'm missing something, so it's a perfect
> example of
> > what we've been discussing back then: there is just no point not
> merging
> > it at any time right ? :-)
>
> It is queued for 2.6.24. I'm not even sure it was originally posted
> in time for the 2.6.23 merge window, but even if it was there was
> a lot of opposition to merging it until fairly recently. In fact,
> I'm sure there are still some wireless developers that are less than
> happy about merging it now.
>
> Anyway, coming soon to a kernel near you...
Allright, thanks. I was mostly trying to figure out where we standed
with this whole idea that driver additions were not necessarily
constrainted by the merge window (which I think is fair to do) and this
looked like a good example to pick since it affects my new laptop :-)
Out of curiosity, what's the main source of opposition ? Since it's
being shipped by distro or built out of tree by most users -anyway-, I
think it's pretty clear that we'd be better off having it merged asap
rather than trying to figure out what random version was included by
users/distros and try to support it, in addition to wider exposure & all
the yadada of being upstream in the first place.
Ben.
On Fri, Sep 28, 2007 at 01:15:23PM +1000, Benjamin Herrenschmidt wrote:
> On Thu, 2007-09-27 at 22:30 -0400, John W. Linville wrote:
> > I'm sure there are still some wireless developers that are less than
> > happy about merging it now.
> Out of curiosity, what's the main source of opposition ? Since it's
> being shipped by distro or built out of tree by most users -anyway-, I
> think it's pretty clear that we'd be better off having it merged asap
> rather than trying to figure out what random version was included by
> users/distros and try to support it, in addition to wider exposure & all
> the yadada of being upstream in the first place.
There have been a lot of spats about how functionality has been
partitioned between the driver and the firmware, issues with how the
driver tries to do things that either ought to be in mac80211 or not
done at all, and some random complaints about ugliness like '#include
"../../../net/mac80211/blah.h"', etc. There is still plenty of work
to be done on this driver, but as you point out we are better-off
with it in the kernel than with it out.
Thanks,
John
--
John W. Linville
[email protected]
On Thu, 27 Sep 2007, Theodore Tso wrote:
> Well, pulling in iwlwifi would require also pulling in the mac80211
> subsystem, so it's not quite that simple (although I'm not sure what's
> holding back that going into the kernel.)
Hi Ted,
sorry? mac80211 is already in the tree, isn't it?
Thanks,
--
Jiri Kosina