* linux-next: procedures
@ 2008-11-12 6:38 Stephen Rothwell
2008-11-18 16:47 ` Randy Dunlap
0 siblings, 1 reply; 3+ messages in thread
From: Stephen Rothwell @ 2008-11-12 6:38 UTC (permalink / raw)
To: linux-next; +Cc: LKML, Andrew Morton, Linus
[-- Attachment #1: Type: text/plain, Size: 2329 bytes --]
Hi all,
I just though I would send out a reminder and clarification of the
procedures around the linux-next tree.
Prerequisites for inclusion of a tree (a branch of a git tree or a quilt
series) in linux-next:
- all its commits/patches have been posted to appropriate mailing
lists
- they have all been reviewed
- the tree has been "unit tested"
- the tree is destined for the next merge window (except during a
merge window when only changes destined for the current merge window
should be there).
So, to be clear, the posting, reviewing and testing should be done
*before* commits/patches are added to linux-next. linux-next is meant to
be for integration testing only i.e. the trees should be as close to what
will be sent to Linus as possible (but may need changes to cope with
changes in other subsystems/architectures).
These trees should be usually based on Linus' tree. In some cases there
will be dependencies and we will deal with these on a case by case basis
(there have not been many so far).
The following will cause a tree to be temporarily dropped from linux-next:
- non-trivial conflicts with Linus' tree
- build failures
- non-obvious conflicts with other trees (this will require
us to come up with some way forward for the trees involved)
- the contact for the tree being unresponsive
Most conflicts will be notified to the contacts for the trees involved
(some really trivial ones will not). Simple conflicts between trees I
will try to fix up (if possible) and carry such resolutions as necessary.
I will not, any more, carry fix up patches to make linux-next build or
boot. If a tree is identified as causing such a problem, it will be
dropped until the problem is fixed.
This is all open to discussion ...
The result I am aiming for is that during a merge window, the trees in
linux-next are sent to Linus verbatim (or with minor conflict fixups) by
their owners and the linux-next ends up basically empty (wrt Linus' tree)
when -rc1 is released (except for a few trees that need to be merged
late). At other times linux-next should be in a state that Andrew can
use it as the basis for an -mm release without to much pain.
--
Cheers,
Stephen Rothwell sfr@canb•auug.org.au
http://www.canb.auug.org.au/~sfr/
[-- Attachment #2: Type: application/pgp-signature, Size: 197 bytes --]
^ permalink raw reply [flat|nested] 3+ messages in thread
* Re: linux-next: procedures
2008-11-12 6:38 linux-next: procedures Stephen Rothwell
@ 2008-11-18 16:47 ` Randy Dunlap
2008-11-18 20:24 ` Stephen Rothwell
0 siblings, 1 reply; 3+ messages in thread
From: Randy Dunlap @ 2008-11-18 16:47 UTC (permalink / raw)
To: Stephen Rothwell; +Cc: linux-next, LKML, Andrew Morton, Linus
On Wed, 12 Nov 2008 17:38:07 +1100 Stephen Rothwell wrote:
...
> The following will cause a tree to be temporarily dropped from linux-next:
> - non-trivial conflicts with Linus' tree
> - build failures
Hi Stephen,
One clarification, please.
Does this mean build failures that _you_ see/experience during tree merging
and not other reported build failures?
Thanks,
~Randy
> - non-obvious conflicts with other trees (this will require
> us to come up with some way forward for the trees involved)
> - the contact for the tree being unresponsive
>
> Most conflicts will be notified to the contacts for the trees involved
> (some really trivial ones will not). Simple conflicts between trees I
> will try to fix up (if possible) and carry such resolutions as necessary.
>
> I will not, any more, carry fix up patches to make linux-next build or
> boot. If a tree is identified as causing such a problem, it will be
> dropped until the problem is fixed.
>
> This is all open to discussion ...
^ permalink raw reply [flat|nested] 3+ messages in thread
* Re: linux-next: procedures
2008-11-18 16:47 ` Randy Dunlap
@ 2008-11-18 20:24 ` Stephen Rothwell
0 siblings, 0 replies; 3+ messages in thread
From: Stephen Rothwell @ 2008-11-18 20:24 UTC (permalink / raw)
To: Randy Dunlap; +Cc: linux-next, LKML, Andrew Morton, Linus
[-- Attachment #1: Type: text/plain, Size: 1446 bytes --]
Hi Randy,
On Tue, 18 Nov 2008 08:47:57 -0800 Randy Dunlap <randy.dunlap@oracle•com> wrote:
>
> > The following will cause a tree to be temporarily dropped from linux-next:
> > - non-trivial conflicts with Linus' tree
> > - build failures
>
> One clarification, please.
>
> Does this mean build failures that _you_ see/experience during tree merging
> and not other reported build failures?
Build failures I see during merging will cause the tree to be dropped
immediately (this forces me to not spend time trying to fix others
problems), other build failures may take a day or two before the
offending tree gets identified and dropped (but hopefully it will be
fixed before that happens).
A tree being dropped during merging does not imply fault - it may well be
a combination of trees that is the problem and I am always open to people
explaining things to me. As long as we come up with a solution that I
can carry along. However, most of the build failures I have seen so far
have been unexpected impact on other architectures/configs (than those
tested by the author), procedures gone wrong or carelessness ...
Trees do get picked up again at the next opportunity i.e. I still
fetch, inspect and attempt to merge "dropped" trees just in case people
don't get around to telling me they are fixed.
--
Cheers,
Stephen Rothwell sfr@canb•auug.org.au
http://www.canb.auug.org.au/~sfr/
[-- Attachment #2: Type: application/pgp-signature, Size: 197 bytes --]
^ permalink raw reply [flat|nested] 3+ messages in thread
end of thread, other threads:[~2008-11-18 20:25 UTC | newest]
Thread overview: 3+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2008-11-12 6:38 linux-next: procedures Stephen Rothwell
2008-11-18 16:47 ` Randy Dunlap
2008-11-18 20:24 ` Stephen Rothwell
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox