BK to GIT migration status for ia64

BK to GIT migration status for ia64

[Luck, Tony]

It's been a hectic month (well, not even a month ... this
is only day 24 of the great migration :)

I've set up a pair of GIT trees on kernel.org for ia64 to
keep track of changes.

1) rsync://rsync.kernel.org/pub/scm/linux/kernel/git/aegl/linux-2.6.git

This is the "to-linus" tree.  Anything commited here is queued to go
into Linus' tree next time he does a pull from me.  Some obviously
correct (to me) patches will go directly into this tree, but most
patches will spend time in the test tree.

2) rsync://rsync.kernel.org/pub/scm/linux/kernel/git/aegl/test-2.6.git

This is the "to-andrew" tree.  Andrew has a script to pull automatically
from here, so changes in this tree will usually end up in the next
"-mm" patch.

I'm not going to play games renaming these trees each release (pauses
to wait for cheering to die down) this didn't really work for me the
way it did for Len (from whom I stole the idea) because my workflow
isn't quite the same.


My BK trees are now offically dead.  Most of the live patches that
were cooking in the "-test" tree have gone into Linus's tree.  The
exception is a dozen or so syscall improvement patches from David
that are still cooking in -mm, they are in the test-2.6.git tree to
finish up, and will move on to Linus after 2.6.12 comes out.
If you see anything that was in a BK tree, and isn't in GIT, then
please let me know.


I'm still working out the details of internal workflow ... trees for
patches to sit in while the testing phase in -mm runs its course. I
don't want to be constrained to have every patch test for the same
period of time ... so I think I still need a small forest of trees from
which to cherry pick into the two public trees.

-Tony

Re: BK to GIT migration status for ia64

[David Mosberger]

>>>>> On Thu, 28 Apr 2005 12:19:49 -0700, "Luck, Tony" <tony.luck@intel.com> said:

  Tony> It's been a hectic month (well, not even a month ... this
  Tony> is only day 24 of the great migration :)

  Tony> I've set up a pair of GIT trees on kernel.org for ia64 to
  Tony> keep track of changes.

  Tony> 1) rsync://rsync.kernel.org/pub/scm/linux/kernel/git/aegl/linux-2.6.git

  Tony> This is the "to-linus" tree.  Anything commited here is queued to go
  Tony> into Linus' tree next time he does a pull from me.  Some obviously
  Tony> correct (to me) patches will go directly into this tree, but most
  Tony> patches will spend time in the test tree.

  Tony> 2) rsync://rsync.kernel.org/pub/scm/linux/kernel/git/aegl/test-2.6.git

  Tony> This is the "to-andrew" tree.  Andrew has a script to pull
  Tony> automatically from here, so changes in this tree will usually
  Tony> end up in the next "-mm" patch.

Looking good so far.  Of course, I'm still waiting to see what happens
when attempting a first non-trivial merge...

  Tony> I'm not going to play games renaming these trees each release (pauses
  Tony> to wait for cheering to die down) this didn't really work for me the
  Tony> way it did for Len (from whom I stole the idea) because my workflow
  Tony> isn't quite the same.

Excellent.  Might have to wait for a _long_ while for the applause to
die down... ;-)

	--david

Re: BK to GIT migration status for ia64

[Keith Owens]

On Thu, 28 Apr 2005 12:19:49 -0700, 
"Luck, Tony" <tony.luck@intel.com> wrote:
>1) rsync://rsync.kernel.org/pub/scm/linux/kernel/git/aegl/linux-2.6.git
>2) rsync://rsync.kernel.org/pub/scm/linux/kernel/git/aegl/test-2.6.git

Neither tree contains a refs or heads directory.  Other git trees like
linux and davem have these directories.  cogito cg-clone and cg-update
expect to find refs and heads.

# ls -l /pub/scm/linux/kernel/git/torvalds/linux-2.6.git
total 16
lrwxrwxrwx    1 torvalds torvalds   17 Apr 27 10:09 HEAD -> refs/heads/master
-rw-r--r--    1 torvalds torvalds   41 Apr 30 15:42 MERGE_HEAD
-rw-rw-r--    1 torvalds torvalds   41 Apr 30 15:42 ORIG_HEAD
drwxr-xr-x  258 torvalds torvalds 4096 Apr 25 08:23 objects
drwxrwxr-x    3 torvalds torvalds 4096 Apr 26 15:23 refs
# ls -l /pub/scm/linux/kernel/git/davem/net-2.6.git
total 1612
lrwxrwxrwx    1 davem davem      17 Apr 28 12:38 HEAD -> refs/heads/master
-rw-r--r--    1 davem davem      41 Apr 17 20:01 MERGE-HEAD
-rw-rw-r--    1 davem davem      41 Apr 27 10:05 MERGE_HEAD
-rw-rw-r--    1 davem davem      41 Apr 27 10:05 ORIG_HEAD
drwxrwxr-x    2 davem davem    4096 Apr 20 16:38 heads
-rw-------    1 davem davem 1620168 Apr 28 12:16 index
drwxr-sr-x  258 davem davem    4096 Apr 21 10:21 objects
drwxrwxr-x    3 davem davem    4096 Apr 26 15:23 refs
# ls -l /pub/scm/linux/kernel/git/aegl/linux-2.6.git
total 12
-rw-rw-r--    1 aegl ftpadmin   41 Apr 27 14:36 HEAD
-rw-rw-r--    1 aegl ftpadmin   38 Apr 29 17:24 description
drwxr-sr-x  258 aegl ftpadmin 4096 Apr 21 10:21 objects
# ls -l /pub/scm/linux/kernel/git/aegl/test-2.6.git/
total 12
-rw-rw-r--    1 aegl ftpadmin   41 Apr 27 21:22 HEAD
-rw-rw-r--    1 aegl ftpadmin   36 Apr 29 17:25 description
drwxr-sr-x  258 aegl ftpadmin 4096 Apr 27 21:39 objects

Re: BK to GIT migration status for ia64

[David Mosberger]

>>>>> On Sun, 01 May 2005 14:19:28 +1000, Keith Owens <kaos@sgi.com> said:

  Keith> On Thu, 28 Apr 2005 12:19:49 -0700, "Luck, Tony"
  Keith> <tony.luck@intel.com> wrote:
  >> 1)
  >> rsync://rsync.kernel.org/pub/scm/linux/kernel/git/aegl/linux-2.6.git
  >> 2)
  >> rsync://rsync.kernel.org/pub/scm/linux/kernel/git/aegl/test-2.6.git

  Keith> Neither tree contains a refs or heads directory.  Other git
  Keith> trees like linux and davem have these directories.  cogito
  Keith> cg-clone and cg-update expect to find refs and heads.

I cloned the test-2.6 tree last Friday (with cogito) without any
problems.  I do see the "refs" and "heads" directories.  Not that it
helps explain your problem, but just an additional data-point.

	--david

Re: BK to GIT migration status for ia64

[Keith Owens]

On Mon, 2 May 2005 04:47:09 -0700, 
David Mosberger <davidm@napali.hpl.hp.com> wrote:
>>>>>> On Sun, 01 May 2005 14:19:28 +1000, Keith Owens <kaos@sgi.com> said:
>
>  Keith> On Thu, 28 Apr 2005 12:19:49 -0700, "Luck, Tony"
>  Keith> <tony.luck@intel.com> wrote:
>  >> 1)
>  >> rsync://rsync.kernel.org/pub/scm/linux/kernel/git/aegl/linux-2.6.git
>  >> 2)
>  >> rsync://rsync.kernel.org/pub/scm/linux/kernel/git/aegl/test-2.6.git
>
>  Keith> Neither tree contains a refs or heads directory.  Other git
>  Keith> trees like linux and davem have these directories.  cogito
>  Keith> cg-clone and cg-update expect to find refs and heads.
>
>I cloned the test-2.6 tree last Friday (with cogito) without any
>problems.  I do see the "refs" and "heads" directories.  Not that it
>helps explain your problem, but just an additional data-point.

Those directories are created locally by the clone process, but they do
not exist in the kernel.org trees.  I'm not sure if they are needed or
not, almost every other git tree on kernel.org has them.

RE: BK to GIT migration status for ia64

[Luck, Tony]

>Neither tree contains a refs or heads directory.  Other git trees like
>linux and davem have these directories.  cogito cg-clone and cg-update
>expect to find refs and heads.

Fixed.

-Tony