On Tue, Apr 19, 2005 at 12:40:44PM -0700, Linus Torvalds wrote:

I'm still working out some performance issues with merges (the actual
"merge" operation itself is very fast, but I've been trying to make the
subsequent "update the working directory tree to the right thing" be much
better).

Ok, if you want some practice with "real" merges, feel free to merge from
the following two trees whenever you are ready:
	kernel.org/pub/scm/linux/kernel/git/gregkh/aoe-2.6.git/
for 11 aoe bugfix patches, and:
	kernel.org/pub/scm/linux/kernel/git/gregkh/driver-2.6.git/
for 13 driver core, sysfs, and debugfs fixes.

The diffstats are:

aoe:
 Documentation/aoe/mkdevs.sh       |    1 
 Documentation/aoe/mkshelf.sh      |    1 
 Documentation/aoe/todo.txt        |   14 ++++
 Documentation/aoe/udev-install.sh |    6 +-
 drivers/block/aoe/aoe.h           |   23 +++++--
 drivers/block/aoe/aoeblk.c        |    5 +
 drivers/block/aoe/aoecmd.c        |  110 ++++++++++++++++++++++----------------
 drivers/block/aoe/aoedev.c        |    8 +-
 drivers/block/aoe/aoenet.c        |    8 +-
 9 files changed, 111 insertions(+), 65 deletions(-)

driver:
 Documentation/kref.txt        |  221 +++++++++++++++++++++++++++++++++++++++++-
 drivers/base/class.c          |    2 
 drivers/base/core.c           |    3 
 drivers/base/firmware_class.c |    3 
 drivers/base/platform.c       |    1 
 drivers/usb/host/hc_crisv10.c |    1 
 fs/partitions/check.c         |    2 
 fs/sysfs/file.c               |   35 ++++++
 include/linux/debugfs.h       |   14 +-
 include/linux/sysfs.h         |    7 +
 lib/kobject.c                 |    7 -
 net/bridge/br_sysfs_if.c      |    2 
 scripts/ver_linux             |    2 
 13 files changed, 290 insertions(+), 10 deletions(-)

No rush, but they should be good test for your merge speeds, as they are
based off of an older HEAD than your current one :)

In other words, I want it to be at the point where people can do

	git pull <repo-address>

and it will "just work", at least for people who don't have any local
changes in their tree. None of this "check out all the files again" crap.

That would be very nice to have.

But how about a plan that we go "live" tomorrow - assuming nobody finds
any problems before that, of course.

That's fine with me.

thanks,

greg k-h