Jody McIntyre wrote:
Do you have any strong feelings as to what we should do with these
patches? My practice has been to send them to akpm for testing in -mm,
then to Linus after 2 weeks. All the fixes are for fairly longstanding
issues AFAICT, so probably not urgent enough to skip that.
What Linus and Andrew do with these patches is ultimately up to them. In
my opinion they could/should go straight into Linus' tree because they
are ...well... fixes and cleanups. There are no new features, no risky
rewrites.
Furthermore, there _is_ a certain urgency IMO. Linux1394.org's stuff and
Linus' tree are already diverging again. So I became a little bit
worried, which is one of the reasons why I put these patches together.
Also note that one group of patches is missing from this collection, as
I already mentioned. I left sbp2's RBC related changes out. (I promise
to produce a patch against the then latest -rc or -rc-mm with these
changes as soon as I have the time; probably at the beginning of
October. Of course I won't mind if somebody else did it...) _This_ kind
of changes belongs into -mm first. (Actually it was already there; made
it even into Linus' tree, but luckily was reverted.)
This is actually another reason why the other patches should better go
into Linus' tree already. We have to have released kernels with an sbp2
driver that actually works. Only a working, released kernel as basis of
reference makes it possible to actually test changes like the RBC
related ones.
Also note: Incarnations of the sbp2 hot-unplug patch, the whole
serialize_io matter, and patch 5/13 (reordering to nodemgr's IRM duties)
have actually received testing by kind readers of linux1394-devel and
linux1394-user. This is certainly a different circle of testers than
those who are running -mm kernels. However it's a very valid way of
testing because there were _fixes_ to test, not new features or code
rearrangements.
(That was a long answer... However, don't take me wrong. I don't want to
push anything and I don't have strong feelings what should happen with
the patches.)
In any case,
they should be sent in by Ben or myself, to avoid confusing people :)
Well, _somebody_ should do it. :-)
And I agree that confusion should be avoided. No need to add more
friction to the process.
--
Stefan Richter
-=====-=-=-= =--= =-=--
http://arcgraph.de/sr/
-
: send the line "unsubscribe linux-scsi" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
