Skip to contentSkip to conversation

GFS2 & DLM merge request

DISCUSSION Steven Whitehouse 3 participants 4 messages Sep 22, 2006

lore
Share & tools
lore
#0 lore raw

Hi,

Linus, I believe that all the outstanding issues raised by Christoph, Jan and others relating to GFS2 and DLM are now settled. Please therefore consider pulling from:

	git://git.kernel.org/pub/scm/linux/kernel/git/steve/gfs2-2.6.git

The diffstat is below. Thanks are due to everybody who has sent us comments, suggestions and patches, particularly the aforementioned reviewers,

Steve.

---------------------------------------------------------------------------------------------

 CREDITS                            |    6 
 Documentation/filesystems/gfs2.txt |   43 
 MAINTAINERS                        |   18 
 fs/Kconfig                         |    2 
 fs/Makefile                        |    2 
 fs/configfs/item.c                 |    2 
 fs/dlm/Kconfig                     |   21 
 fs/dlm/Makefile                    |   19 
 fs/dlm/ast.c                       |  173 +
 fs/dlm/ast.h                       |   26 
 fs/dlm/config.c                    |  789 +++++++
 fs/dlm/config.h                    |   42 
 fs/dlm/debug_fs.c                  |  387 +++
 fs/dlm/dir.c                       |  423 ++++
 fs/dlm/dir.h                       |   30 
 fs/dlm/dlm_internal.h              |  543 +++++
 fs/dlm/lock.c                      | 3871 +++++++++++++++++++++++++++++++++++++
 fs/dlm/lock.h                      |   62 
 fs/dlm/lockspace.c                 |  717 ++++++
 fs/dlm/lockspace.h                 |   25 
 fs/dlm/lowcomms.c                  | 1238 +++++++++++
 fs/dlm/lowcomms.h                  |   26 
 fs/dlm/lvb_table.h                 |   18 
 fs/dlm/main.c                      |   97 
 fs/dlm/member.c                    |  327 +++
 fs/dlm/member.h                    |   24 
 fs/dlm/memory.c                    |  116 +
 fs/dlm/memory.h                    |   29 
 fs/dlm/midcomms.c                  |  140 +
 fs/dlm/midcomms.h                  |   21 
 fs/dlm/rcom.c                      |  472 ++++
 fs/dlm/rcom.h                      |   24 
 fs/dlm/recover.c                   |  765 +++++++
 fs/dlm/recover.h                   |   34 
 fs/dlm/recoverd.c                  |  290 ++
 fs/dlm/recoverd.h                  |   24 
 fs/dlm/requestqueue.c              |  184 +
 fs/dlm/requestqueue.h              |   22 
 fs/dlm/user.c                      |  788 +++++++
 fs/dlm/user.h                      |   16 
 fs/dlm/util.c                      |  161 +
 fs/dlm/util.h                      |   22 
 fs/gfs2/Kconfig                    |   44 
 fs/gfs2/Makefile                   |   10 
 fs/gfs2/acl.c                      |  309 ++
 fs/gfs2/acl.h                      |   39 
 fs/gfs2/bmap.c                     | 1220 +++++++++++
 fs/gfs2/bmap.h                     |   31 
 fs/gfs2/daemon.c                   |  196 +
 fs/gfs2/daemon.h                   |   19 
 fs/gfs2/dir.c                      | 1962 ++++++++++++++++++
 fs/gfs2/dir.h                      |   79 
 fs/gfs2/eaops.c                    |  230 ++
 fs/gfs2/eaops.h                    |   30 
 fs/gfs2/eattr.c                    | 1501 ++++++++++++++
 fs/gfs2/eattr.h                    |  100 
 fs/gfs2/gfs2.h                     |   31 
 fs/gfs2/glock.c                    | 2231 +++++++++++++++++++++
 fs/gfs2/glock.h                    |  153 +
 fs/gfs2/glops.c                    |  564 +++++
 fs/gfs2/glops.h                    |   25 
 fs/gfs2/incore.h                   |  634 ++++++
 fs/gfs2/inode.c                    | 1340 ++++++++++++
 fs/gfs2/inode.h                    |   56 
 fs/gfs2/lm.c                       |  217 ++
 fs/gfs2/lm.h                       |   42 
 fs/gfs2/locking.c                  |  184 +
 fs/gfs2/locking/dlm/Makefile       |    3 
 fs/gfs2/locking/dlm/lock.c         |  524 +++++
 fs/gfs2/locking/dlm/lock_dlm.h     |  187 +
 fs/gfs2/locking/dlm/main.c         |   64 
 fs/gfs2/locking/dlm/mount.c        |  255 ++
 fs/gfs2/locking/dlm/plock.c        |  301 ++
 fs/gfs2/locking/dlm/sysfs.c        |  226 ++
 fs/gfs2/locking/dlm/thread.c       |  359 +++
 fs/gfs2/locking/nolock/Makefile    |    3 
 fs/gfs2/locking/nolock/main.c      |  246 ++
 fs/gfs2/log.c                      |  578 +++++
 fs/gfs2/log.h                      |   65 
 fs/gfs2/lops.c                     |  809 +++++++
 fs/gfs2/lops.h                     |   99 
 fs/gfs2/main.c                     |  150 +
 fs/gfs2/meta_io.c                  |  753 +++++++
 fs/gfs2/meta_io.h                  |   78 
 fs/gfs2/mount.c                    |  214 ++
 fs/gfs2/mount.h                    |   17 
 fs/gfs2/ondisk.c                   |  308 ++
 fs/gfs2/ops_address.c              |  782 +++++++
 fs/gfs2/ops_address.h              |   22 
 fs/gfs2/ops_dentry.c               |  119 +
 fs/gfs2/ops_dentry.h               |   17 
 fs/gfs2/ops_export.c               |  298 ++
 fs/gfs2/ops_export.h               |   22 
 fs/gfs2/ops_file.c                 |  642 ++++++
 fs/gfs2/ops_file.h                 |   24 
 fs/gfs2/ops_fstype.c               |  943 +++++++++
 fs/gfs2/ops_fstype.h               |   18 
 fs/gfs2/ops_inode.c                | 1151 +++++++++++
 fs/gfs2/ops_inode.h                |   20 
 fs/gfs2/ops_super.c                |  468 ++++
 fs/gfs2/ops_super.h                |   17 
 fs/gfs2/ops_vm.c                   |  184 +
 fs/gfs2/ops_vm.h                   |   18 
 fs/gfs2/quota.c                    | 1226 +++++++++++
 fs/gfs2/quota.h                    |   35 
 fs/gfs2/recovery.c                 |  570 +++++
 fs/gfs2/recovery.h                 |   34 
 fs/gfs2/rgrp.c                     | 1513 ++++++++++++++
 fs/gfs2/rgrp.h                     |   69 
 fs/gfs2/super.c                    |  977 +++++++++
 fs/gfs2/super.h                    |   54 
 fs/gfs2/sys.c                      |  583 +++++
 fs/gfs2/sys.h                      |   27 
 fs/gfs2/trans.c                    |  184 +
 fs/gfs2/trans.h                    |   39 
 fs/gfs2/util.c                     |  245 ++
 fs/gfs2/util.h                     |  170 +
 include/linux/Kbuild               |   39 
 include/linux/dlm.h                |  302 ++
 include/linux/dlm_device.h         |   86 
 include/linux/fs.h                 |    3 
 include/linux/gfs2_ondisk.h        |  443 ++++
 include/linux/iflags.h             |  102 
 include/linux/kernel.h             |    1 
 include/linux/lm_interface.h       |  273 ++
 include/linux/lock_dlm_plock.h     |   41 
 mm/filemap.c                       |    3 
 mm/readahead.c                     |    1 
 128 files changed, 40266 insertions(+), 24 deletions(-)
#1 lore raw
Quoted 5 lines · Steven Whitehouse

Hi,

Linus, I believe that all the outstanding issues raised by Christoph, Jan and others relating to GFS2 and DLM are now settled. Please therefore consider pulling from:

Clear NACK from me. The mails I replied to where only the mails Jan started nitpicking and I looked a little deeper at what he already quoted, maybe 5% of the codebase. Given the horrors I found there I'm pretty sure there will be quite a few more. And given the amount of junk Andrew plans to push to Linus for 2.6.19 I'll be pretty busy to look at that, aswell as looking at things I promised David for month now, so I'm a little busy.

Quoted 5 lines
 include/linux/fs.h                 |    3 
 include/linux/iflags.h             |  102 
 include/linux/kernel.h             |    1 
 mm/filemap.c                       |    3 
 mm/readahead.c                     |    1 

And while we're at it, please don't push core change as part of a subsystem tree ever. They should go into clearly marked and separately patches via -mm.

#2 lore raw

Hi,

Quoted 7 lines · Christoph Hellwig
Quoted 5 lines · Steven Whitehouse

Hi,

Linus, I believe that all the outstanding issues raised by Christoph, Jan and others relating to GFS2 and DLM are now settled. Please therefore consider pulling from:

Quoted 7 lines

Clear NACK from me. The mails I replied to where only the mails Jan started nitpicking and I looked a little deeper at what he already quoted, maybe 5% of the codebase. Given the horrors I found there I'm pretty sure there will be quite a few more. And given the amount of junk Andrew plans to push to Linus for 2.6.19 I'll be pretty busy to look at that, aswell as looking at things I promised David for month now, so I'm a little busy.

Well I hope that we don't have too many more "horrors" left in the code. It has after all been out for review many times in the last year and I hope we've not done too bad a job of incorporating all the suggested changes so far.

When I spoke to you at OLS you did promise me that you'd look at the code in more detail and that was more than a month ago now. I know there is a lot else to be reviewed as well as GFS2 and we need to wait our place in the queue as it were, but it would be helpful to know where that place is, so we can then try our best to work with the process.

Quoted 9 lines
Quoted 5 lines
 include/linux/fs.h                 |    3 
 include/linux/iflags.h             |  102 
 include/linux/kernel.h             |    1 
 mm/filemap.c                       |    3 
 mm/readahead.c                     |    1 

And while we're at it, please don't push core change as part of a subsystem tree ever. They should go into clearly marked and separately patches via -mm.

Ok, I'm quite happy to do that. They are pretty minor changes, but the usual argument is that such changes are not acceptable until there is something in the kernel that makes use of them, and they have to happen one way around or the other,

Steve.

#3 lore raw
Quoted 6 lines · On Fri, 22 Sep 2006 17:02:24 +0100 Steven Whitehouse <[email protected]>…
Quoted 6 lines
Quoted 5 lines
 include/linux/fs.h                 |    3 
 include/linux/iflags.h             |  102 
 include/linux/kernel.h             |    1 
 mm/filemap.c                       |    3 
 mm/readahead.c                     |    1 
Quoted 8 lines
Quoted 3 lines

And while we're at it, please don't push core change as part of a subsystem tree ever. They should go into clearly marked and separately patches via -mm.

Ok, I'm quite happy to do that. They are pretty minor changes, but the usual argument is that such changes are not acceptable until there is something in the kernel that makes use of them, and they have to happen one way around or the other,

There's no 100% comfortable solution to this. Steven presumably needs those changes present in his tree so he can test.

What I usually suggest is that the developer get the core changes reviewed on-list at an early stage and then explicitly flag their presence in the covering description when sending the please-pull email.