Thread (34 messages) 34 messages, 7 authors, 2012-05-23

Re: [tip:perf/uprobes] uprobes, mm, x86: Add the ability to install and remove uprobes breakpoints

From: Andrew Morton <akpm@linux-foundation.org>
Date: 2012-05-22 02:25:52
Also in: lkml

On Tue, 22 May 2012 11:16:18 +1000 Stephen Rothwell [off-list ref] wrote:
Hi Andrew,

On Mon, 21 May 2012 15:13:23 -0700 Andrew Morton [off-list ref] wrote:
quoted
On Mon, 21 May 2012 15:00:28 -0700
Linus Torvalds [off-list ref] wrote:
quoted
On Mon, May 21, 2012 at 2:37 PM, Andrew Morton
[off-list ref] wrote:
quoted
hm, we seem to have conflicting commits between mainline and linux-next.
During the merge window. __Again. __Nobody knows why this happens.
I didn't have my trivial cleanup branches in linux-next, I'm afraid.
Well, it's a broader issue than that.  I often see a large number of
rejects when syncing mainline with linux-next during the merge window. 
Right now:
Some of that is because your patch series is based on the end of
linux-next and part way through the merge window only some of that has
been merged by Linus.  Also some of it gets rebased before Linus is asked
to pull (a real pain) - there hasn't been much of that (yet) this merge
window (but its early days :-().  Also, sometimes Linus' merge
resolutions are different to mine.

I have been meaning to talk to you about basing the majority of your
patch series on Linus' tree.  This would give it mush greater stability
and would make the merge resolution my problem (and Linus', of course).
Confused.  None of those conflicts have anything to do with the -mm
patches: the only trees involved there are mainline and
trees-in-next-other-than-mm.
There will be bits that may need to be based on other work in linux-next,
but I suspect that it is not very much.
Well, there are a number of reasons why I base off linux-next.  To see
whether others have merged patches which I have merged (and, sometimes,
missed later fixes to them).  Explicit fixes against -next material. 
To get visibility into upcoming merge problems.  And so that I and
others test -next too.

Basing -mm on next is never a problem (for me).  What is a problem is
the mess which happens when people merge things into mainline which are
(I assume) either slightly different from what they merged in -next or
which never were in -next at all.

That's guessing - it's a long time since I sat down and worked out exactly
what is causing this.

--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@kvack.org.  For more info on Linux MM,
see: http://www.linux-mm.org/ .
Fight unfair telecom internet charges in Canada: sign http://stopthemeter.ca/
Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>
Keyboard shortcuts
hback out one level
jnext message in thread
kprevious message in thread
ldrill in
Escclose help / fold thread tree
?toggle this help