ack
Tiger Yang wrote:
We used to return positive EAGAIN to indicate a retry action
is needed in dlm_begin_reco_handler(). Now we return negative
-EAGAIN to erase the confusion caused by this error code.
Signed-off-by: Tiger Yang tiger.y...@oracle.com
---
fs/ocfs2/dlm/dlmrecovery.c | 18
Joel Becker wrote:
It's yucky, but I know what you're doing. I hope we're going to
be able to re-enable quotas?
I saved the yuck for last. ;)
___
Ocfs2-devel mailing list
Ocfs2-devel@oss.oracle.com
Signed-off-by: Sunil Mushran sunil.mush...@oracle.com
comments inlined
Tiger Yang wrote:
We used to return positive EAGAIN to indicate a retry action
is needed in dlm_begin_reco_handler(). Now we use negative -EAGAIN
to erase the confusion caused by this error code.
Signed-off-by: Tiger
panic /sys/kernel/config/cluster/clustername/fence_method
Signed-off-by: Sunil Mushran sunil.mush...@oracle.com
---
fs/ocfs2/cluster/nodemanager.c | 50
fs/ocfs2/cluster/nodemanager.h |7 +
fs/ocfs2/cluster/quorum.c | 14 +-
3 files
Mainline commit e4ad08fe64afca4ef79ecc4c624e6e871688da0d added macro
BDI_CAP_NO_ACCT_AND_WRITEBACK.
Signed-off-by: Sunil Mushran sunil.mush...@oracle.com
---
Makefile |3 ++-
configure.in |5 +
kapi-compat/include/bdi_define.h |6
Mainline commit e0bf68ddec4f4f90e5871404be4f1854c17f3120 added bdi_init()
and bdi_destroy(). Patch adds empty inlines to allow building with EL5.
Signed-off-by: Sunil Mushran sunil.mush...@oracle.com
---
Makefile |3 ++-
configure.in |5 +
kapi
Mainline commit 20c2df83d25c6a95affe6157a4c9cac4cf5ffaac removed the
dtor arg from kmem_cache_create(). Patch adds a kapi macro that adds
a dtor arg when building against EL5 kernels.
Signed-off-by: Sunil Mushran sunil.mush...@oracle.com
---
Config.make.in |1 +
Makefile
Commit 19582597eec08f61b6f5794c15f6f2c9821a67d9 added kapi-default.h in
fs/ocfs2. This patch adds a similar named file in fs/ocfs2/cluster and
fs/ocfs2/dlm to allow building against the mainline without kapi-compat
headers.
Signed-off-by: Sunil Mushran sunil.mush...@oracle.com
---
fs/ocfs2
-by: Sunil Mushran sunil.mush...@oracle.com
---
Makefile |3 ++-
configure.in |5 +
fs/ocfs2/file.c |6 +++---
fs/ocfs2/kapi-default.h |6 ++
kapi-compat
against current and EL5 kernels.
Signed-off-by: Sunil Mushran sunil.mush...@oracle.com
---
Makefile |3 ++-
configure.in |5 +
fs/ocfs2/kapi-default.h|9 +
fs/ocfs2/refcounttree.c| 18
Mainline commit 9a1c3542768b5a58e45a9216921cd10a3bae1205 added arg fmode_t
to blkdev_put(). Patch allows building against EL5 by ignoring that argument.
Signed-off-by: Sunil Mushran sunil.mush...@oracle.com
---
Makefile |3 ++-
configure.in |5
Mainline commit d23a147bb6e8d467e8df73b6589888717da3b9ce exported the symbol
should_remove_suid(). Patch adds compat_should_remove_suid.c that includes
the same function that gets built when building with EL5.
Signed-off-by: Sunil Mushran sunil.mush...@oracle.com
---
Config.make.in
Mainline commit e6bd07aee739566803425acdbf5cdb29919164e1 changed a semaphore
(su_sem)
into a mutex (su_mutex) in struct configfs_subsystem. Patch adds macros to allow
building with EL5.
Signed-off-by: Sunil Mushran sunil.mush...@oracle.com
---
Makefile|3
Mainline commit 91cf45f02af5c871251165d000c3f42a2a0b0552 added helper
kernel_sock_shutdown().
Signed-off-by: Sunil Mushran sunil.mush...@oracle.com
---
Makefile|3 ++-
configure.in|5 +
kapi-compat/include/sock_shutdown.h | 17
Mainline commit f419a2e3b64def707e1384ee38abb77f99af5f6d renamed permission()
to inode_permission(). Patch uses an inline to redirect inode_permission() to
permission() when building with EL5.
Signed-off-by: Sunil Mushran sunil.mush...@oracle.com
---
Makefile |3
Mainline commit 6712ecf8f648118c3363c142196418f89a510b90 dropped the 'size'
argument from bi_end_io_t. Also changed the return from int to void. Patch
uses the older prototype when building with EL5 kernels.
Signed-off-by: Sunil Mushran sunil.mush...@oracle.com
---
Config.make.in
Mainline commit 54cb8821de07f2ffcd28c380ce9b93d5784b40d7 introduced
vmops-fault()
to replace vmops-populate() and vmops-nopage(). Patch adds nopage() and uses
it
when building against EL5.
Signed-off-by: Sunil Mushran sunil.mush...@oracle.com
---
Config.make.in|1 +
configure.in
The next set. All patches prior to this have been pushed to the repo.
___
Ocfs2-devel mailing list
Ocfs2-devel@oss.oracle.com
http://oss.oracle.com/mailman/listinfo/ocfs2-devel
Commit dbc80b8f391820b719db97004219327c8cf050bc added KAPI_INIT_WORK
macros to fs/ocfs2/kapi-default.h. This patch adds the same to the
same header in dlm and cluster sub-directories.
Signed-off-by: Sunil Mushran sunil.mush...@oracle.com
---
fs/ocfs2/cluster/kapi-default.h |6 ++
fs
Joel Becker wrote:
Rather than include kapi-compat/include/workqueue.h at all
times, I'd really like to only include it when necessary. That way you
can build the driver against mainline even when there is no kapi-compat
directory.
Can't we have an fs/ocfs2/kapi-default.h that
Joel Becker wrote:
Again you're including the kapi-compat header unconditionally.
Only include it when it is necessary.
Ok.
+#ifdef NO_DLMCONSTANTS_HEADER
+#include linux/types.h
+#include linux/dlm.h
Then maybe change this to
#ifndef DLM_LOCK_NL
# include
Mainline commit 0f7fc9e4d03987fe29f6dd4aa67e4c56eb7ecb05 added
struct path into struct file.
Signed-off-by: Sunil Mushran sunil.mush...@oracle.com
---
Config.make.in |1 +
Makefile|3 ++-
configure.in|6 ++
fs/ocfs2/Makefile
Add kapi-default.h to allow us to build ocfs2 1.6 against mainline
without the kapi-compat headers.
Signed-off-by: Sunil Mushran sunil.mush...@oracle.com
---
fs/ocfs2/Makefile |2 +-
fs/ocfs2/kapi-default.h | 26 ++
fs/ocfs2/ocfs2.h|2 ++
3 files
Added kapi-default.h to enable us to not have to include kapi-compat
headers when building against mainline.
Sunil
___
Ocfs2-devel mailing list
Ocfs2-devel@oss.oracle.com
http://oss.oracle.com/mailman/listinfo/ocfs2-devel
Mainline commit 2b47c3611de05c585e2d81204f6c7e3e255a3461 changed type of
file-f_version from unsigned long to u64. This patch allows building with
kernels having either definition. Build fails if the type is something else.
Signed-off-by: Sunil Mushran sunil.mush...@oracle.com
---
Config.make.in
Mainline commit d69a3ad6a0e47b2aa9b2b2ddfd385752132a4d34 split
lock mode and flag constants from dlm.h into a sharable header,
dlmconstants.h.
This patch adds a header with the same name that merely includes
dlm.h.
Signed-off-by: Sunil Mushran sunil.mush...@oracle.com
---
Config.make.in
Mainline commit 52bad64d95bd89e08c49ec5a071fa6dcbe5a1a9c introduced
struct delayed_work to separate delayable and non-delayable work.
Signed-off-by: Sunil Mushran sunil.mush...@oracle.com
---
Config.make.in |2 +
Makefile|3 +-
configure.in
Mainline commit 027445c37282bc1ed26add45e573ad2d3e4860a5 vectorized fileops
aio_read() and aio_write().
Patch provides support for the non-vectorized fileops (in EL5) by hand crafting
the iovec with the provided buf and buflen.
Signed-off-by: Sunil Mushran sunil.mush...@oracle.com
Mainline commit 0e03036c97b70b2602f7dedaa3a223ed7563c2c9 added enum FS_OCFS2.
Signed-off-by: Sunil Mushran sunil.mush...@oracle.com
---
Makefile |3 ++-
configure.in |5 +
kapi-compat/include/sysctl.h |9 +
3 files changed, 16
The next set. The previous 3 sets have been pushed to the git repo.
___
Ocfs2-devel mailing list
Ocfs2-devel@oss.oracle.com
http://oss.oracle.com/mailman/listinfo/ocfs2-devel
Mainline commit 0b4d414714f0d2f922d39424b0c5c82ad900a381 removed the
insert_at_head
argument from register_sysctl_table(). Patch allows building with older kernels.
Signed-off-by: Sunil Mushran sunil.mush...@oracle.com
---
Makefile |3 ++-
configure.in
Mainline commit 86a264abe542cfececb4df129bc45a0338d8cdb9 added helper
current_fsuid().
An earlier commit grouped security credentials in struct task in struct cred.
The patch ensures that if the above helper is not found, that struct cred is
missing. If not, build fails.
Signed-off-by: Sunil
the older prototype.
Signed-off-by: Sunil Mushran sunil.mush...@oracle.com
---
Config.make.in|1 +
configure.in | 13 +
fs/ocfs2/Makefile |4
fs/ocfs2/file.c |4
fs/ocfs2/file.h |4
5 files changed, 26 insertions(+), 0 deletions(-)
diff --git
Mainline commit a16877ca9cec211708a161057a7cbfbf2cbc3a53 added helper
mandatory_lock().
Signed-off-by: Sunil Mushran sunil.mush...@oracle.com
---
Makefile |3 ++-
configure.in |4
kapi-compat/include/mandatory_lock.h | 25
-by: Sunil Mushran sunil.mush...@oracle.com
---
Makefile |3 ++-
configure.in |5 +
kapi-compat/include/relatime.h |6 ++
3 files changed, 13 insertions(+), 1 deletions(-)
create mode 100644 kapi-compat/include/relatime.h
diff --git a/Makefile
Mainline commit f0f37e2f77731b3473fa6bd5ee53255d9a9cdb40 const-ified
vm_area_struct-vm_ops.
Signed-off-by: Sunil Mushran sunil.mush...@oracle.com
---
Config.make.in|1 +
configure.in |5 +
fs/ocfs2/Makefile |4
fs/ocfs2/mmap.c |4
4 files changed, 14
Mainline commit ee9b6d61a2a43c5952eb43283f8db284a4e70b8a const-ified sb-s_op.
Signed-off-by: Sunil Mushran sunil.mush...@oracle.com
---
Config.make.in|1 +
configure.in |5 +
fs/ocfs2/Makefile |4
fs/ocfs2/dlm/Makefile |4
fs/ocfs2/dlm/dlmfs.c
Mainline commit 15ad7cdcfd76450d4beebc789ec646664238184d const-ified
seq_file-op.
Signed-off-by: Sunil Mushran sunil.mush...@oracle.com
---
Config.make.in |1 +
configure.in |5 +
fs/ocfs2/Makefile |4
fs/ocfs2/dlmglue.c |4
4 files changed, 14
Mainline commit e56980d451904b623573ef4966cbab768e433c79 const-ified
dentry-d_op.
Signed-off-by: Sunil Mushran sunil.mush...@oracle.com
---
Config.make.in|1 +
configure.in |6 ++
fs/ocfs2/Makefile |4
fs/ocfs2/dcache.c |4
fs/ocfs2/dcache.h |4
5
Mainline commit 39655164405940d4818224a085e35420e2f97aed const-ified
sb-s_export_op.
Signed-off-by: Sunil Mushran sunil.mush...@oracle.com
---
Config.make.in|1 +
configure.in |5 +
fs/ocfs2/Makefile |4
fs/ocfs2/export.c |4
fs/ocfs2/export.h |4
The third set of patches that deal with missing const modifier in various
places in the el5 kernel.
___
Ocfs2-devel mailing list
Ocfs2-devel@oss.oracle.com
http://oss.oracle.com/mailman/listinfo/ocfs2-devel
Mainline commit c5ef1c42c51b1b5b4a401a6517bdda30933ddbaf const-ified
inode-i_op.
Signed-off-by: Sunil Mushran sunil.mush...@oracle.com
---
Config.make.in|1 +
configure.in |5 +
fs/ocfs2/Makefile |4
fs/ocfs2/dlm/Makefile |4
fs/ocfs2/dlm
Mainline commit 15ad7cdcfd76450d4beebc789ec646664238184d added the const
modifier to struct seq_operations * in seq_open().
Signed-off-by: Sunil Mushran sunil.mush...@oracle.com
---
Config.make.in |1 +
configure.in|5 +
fs/ocfs2/cluster/Makefile |4
So the ocfs2 1.6 repository is up. It is a copy of the fs in the mainline.
I have some 70-80 patches that will make it run atop el5 u4. The patches
will be posted in smaller batches so as to make it easier to review.
The first set handles the makefiles, etc.
Sunil
Teach the makefile to handle stackglue and friends.
Signed-off-by: Sunil Mushran sunil.mush...@oracle.com
---
fs/ocfs2/Makefile | 115 -
1 files changed, 34 insertions(+), 81 deletions(-)
diff --git a/fs/ocfs2/Makefile b/fs/ocfs2/Makefile
Signed-off-by: Sunil Mushran sunil.mush...@oracle.com
---
fs/ocfs2/Makefile | 212 ++---
fs/ocfs2/cluster/Makefile | 89 ++-
fs/ocfs2/dlm/Makefile | 105 +-
3 files changed, 347 insertions(+), 59 deletions
Signed-off-by: Sunil Mushran sunil.mush...@oracle.com
---
.gitignore | 43 +++
fs/ocfs2/.gitignore | 35 +++
fs/ocfs2/cluster/.gitignore | 35 +++
fs/ocfs2/dlm
Mainline commit 5b04aa3a64f854244bc40a6f528176ed50b5c4f6 added
do_sync_mapping_range(). EL5 has do_sync_file_range(). This patch
wires do_sync_mapping_range() into do_sync_file_range().
Build fails if both symbols are not found.
Signed-off-by: Sunil Mushran sunil.mush...@oracle.com
---
Makefile
Mainline commit d69a3ad6a0e47b2aa9b2b2ddfd385752132a4d34 split
lock mode and flag constants from dlm.h into a sharable header,
dlmconstants.h.
This patch adds a header with the same name that merely includes
dlm.h.
Signed-off-by: Sunil Mushran sunil.mush...@oracle.com
---
Config.make.in
Mainline commit 2b47c3611de05c585e2d81204f6c7e3e255a3461 changed type of
file-f_version from unsigned long to u64. This patch allows building with
kernels having either definition. Build fails if the type is something else.
Signed-off-by: Sunil Mushran sunil.mush...@oracle.com
---
Config.make.in
Mainline commit 52bad64d95bd89e08c49ec5a071fa6dcbe5a1a9c introduced
struct delayed_work to separate delayable and non-delayable work.
Signed-off-by: Sunil Mushran sunil.mush...@oracle.com
---
Config.make.in |2 +
Makefile|3 +-
configure.in
The second set of patches.
___
Ocfs2-devel mailing list
Ocfs2-devel@oss.oracle.com
http://oss.oracle.com/mailman/listinfo/ocfs2-devel
Mainline commit 7af5729474b5b8ad385adadab78d6e723e7655a3 added a helper
to obtain the pid number.
Signed-off-by: Sunil Mushran sunil.mush...@oracle.com
---
Makefile |3 ++-
configure.in |5 +
kapi-compat/include/task_pid_nr.h | 12
Signed-off-by: Sunil Mushran sunil.mush...@oracle.com
---
fs/ocfs2/Makefile |2 ++
1 files changed, 2 insertions(+), 0 deletions(-)
diff --git a/fs/ocfs2/Makefile b/fs/ocfs2/Makefile
index e997fe4..efa7f57 100644
--- a/fs/ocfs2/Makefile
+++ b/fs/ocfs2/Makefile
@@ -34,6 +34,8 @@ EXTRA_CFLAGS
Mainline commit 44f564a4bf6ac70f2a84806203045cf515bc9367 USHORT_MAX.
Signed-off-by: Sunil Mushran sunil.mush...@oracle.com
---
Makefile|3 ++-
configure.in|5 +
kapi-compat/include/ushortmax.h |6 ++
3 files changed, 13 insertions
Mainline commit 0f7fc9e4d03987fe29f6dd4aa67e4c56eb7ecb05 added
struct path into struct file.
Signed-off-by: Sunil Mushran sunil.mush...@oracle.com
---
Config.make.in |1 +
Makefile|3 ++-
configure.in|6 ++
fs/ocfs2/Makefile
Signed-off-by: Sunil Mushran sunil.mush...@oracle.com
---
Documentation/filesystems/ocfs2.txt |6 +++---
1 files changed, 3 insertions(+), 3 deletions(-)
diff --git a/Documentation/filesystems/ocfs2.txt
b/Documentation/filesystems/ocfs2.txt
index c2a0871..c58b9f5 100644
--- a/Documentation
Jamie Lokier wrote:
It might be better to have a flag FIEMAP_EXTENT_NOT_SHARED.
Apps which need to know the data is not shared can't safely check
!(flags FIEMAP_EXTENT_SHARED) because of old kernels which don't set
the flag.
I'm probably not understanding your point. Won't the older kernels
Patch adds flag FIEMAP_EXTENT_SHARED to denote extents that are
shared with other inodes.
Signed-off-by: Sunil Mushran sunil.mush...@oracle.com
---
include/linux/fiemap.h |2 ++
1 files changed, 2 insertions(+), 0 deletions(-)
diff --git a/include/linux/fiemap.h b/include/linux/fiemap.h
Adds FIEMAP_EXTENT_SHARED flag to refcounted extents.
Signed-off-by: Sunil Mushran sunil.mush...@oracle.com
---
fs/ocfs2/extent_map.c |2 ++
1 files changed, 2 insertions(+), 0 deletions(-)
diff --git a/fs/ocfs2/extent_map.c b/fs/ocfs2/extent_map.c
index 843db64..cdce5f8 100644
--- a/fs
This adds a FIEMAP extent flag, FIEMAP_EXTENT_SHARED, to indicate
extents that are shared by multiple files. The second patch makes
use of the same in the OCFS2 file system to show such extents in
the reflinked files.
Thanks
Sunil
___
Ocfs2-devel
Appears the volumes are not mounted with the datavolume mount option.
Refer to the FAQ for more in this.
On May 4, 2009, at 7:37 AM, keyur patel keyur.pa...@oracle.com wrote:
Hello All,
I have installed Oracle Cluster Manager on linux x86-64 nit. I am
using ocfs file system for quorum
Adds FIEMAP_EXTENT_SHARED flag to refcounted extents.
Signed-off-by: Sunil Mushran sunil.mush...@oracle.com
---
fs/ocfs2/extent_map.c |2 ++
1 files changed, 2 insertions(+), 0 deletions(-)
diff --git a/fs/ocfs2/extent_map.c b/fs/ocfs2/extent_map.c
index 843db64..cdce5f8 100644
--- a/fs
Patch adds flag FIEMAP_EXTENT_SHARED to denote extents that are
shared with other inodes.
Signed-off-by: Sunil Mushran sunil.mush...@oracle.com
---
include/linux/fiemap.h |2 ++
1 files changed, 2 insertions(+), 0 deletions(-)
diff --git a/include/linux/fiemap.h b/include/linux/fiemap.h
/cvsmirror/opendlm/docs/dlmbook_final.pdf
Sunil
Coly Li wrote:
Sunil Mushran Wrote:
BAST:
[368.807757] (2572,dlm_astd,0):ocfs2_blocking_ast:1025 BAST fired for
lockres M085e02, blocking 5, level 3 type Meta
[368.807767] (2571,ocfs2dc,0):ocfs2_process_blocked_lock:3839
lockres-l_action = OCFS2_AST_DOWNCONVERT;
3231 lockres-l_requested = new_level;
3232 lockres_or_flags(lockres, OCFS2_LOCK_BUSY);
3233 return lockres_set_pending(lockres);
3234 }
I am trying to understand what you did now :-)
Sunil Mushran Wrote:
So originally my
To use pacemaker today, you will need to install sles11. The pacemaker
bits are being pushed to ocfs2-tools.git. But the git repo is behind
sles11 atm.
On Jun 15, 2009, at 8:52 AM, Jonathan jdccde...@gmail.com wrote:
I have seen many references online to being able to use OCFS2 with
):ocfs2_locking_ast:1069 lock
M085e02, action 3, unlock 0
Upconvert AST:
[369.208048] (2572,dlm_astd,0):ocfs2_locking_ast:1069 lock
M085e02, action 2, unlock 0
Coly Li wrote:
Sunil Mushran Wrote:
The full trace is available here.
http
is consistently 200 usecs (??) after the
DOWNCONVERT.
Coli, Can you map the pids to the process names.
Sunil
Coly Li wrote:
Sunil Mushran Wrote:
Thanks. Can you email me the _full_ trace log.
Hi Sunil,
Here is the full trace log from boot to shutdown, on both nodes.
On node x1, I run
a bit of tlc.
Please re-run with the the following patch.
http://oss.oracle.com/~smushran/0001-ocfs2-Patch-to-debug-hang-in-dlmglue-when-running-d.patch
Coly Li wrote:
Sunil Mushran Wrote:
Can you send me the raw trace output.
Here is the compressed trace info.
Thanks
Signed-off-by: Sunil Mushran sunil.mush...@oracle.com
Tao Ma wrote:
In ocfs2_file_aio_write, we will prevent direct io if
we find that we are appending(changing i_size) and call
generic_file_aio_write_nolock. But actually O_DIRECT flag
is there and this function will call
Maybe so. But I see no harm leaving the check in.
On Sep 9, 2009, at 9:58 AM, Wengang Wang wen.gang.w...@oracle.com
wrote:
there is check on the inode whether it is a DIR in ocfs2_link().
the same check is done in vfs_link already, so remove it from
ocfs2_link().
Signed-off-by: Wengang
Can you send me the raw trace output.
Coly Li wrote:
Sunil Mushran Wrote:
You will have to trace thru process_blocked_lock() to make sense of this.
I try to trace process_blocked_lock(), the result is quite complex. I attach
the
modified fs/ocfs2 code (ocfs2-trace.tar.bz2
Bug introduced by mainline commit e7432675f8ca868a4af365759a8d4c3779a3d922
The bug causes ocfs2_write_begin_nolock() to oops when len=0.
Signed-off-by: Sunil Mushran sunil.mush...@oracle.com
---
fs/ocfs2/aops.c |4 ++--
1 files changed, 2 insertions(+), 2 deletions(-)
diff --git a/fs/ocfs2
Signed-off-by: Sunil Mushran sunil.mush...@oracle.com
Wengang Wang wrote:
when we check/modify lockres-purge, we should with the protection of
lockres-spinlock.
in dlm_purge_lockres(), the checking/modifying is not with the protectin.
this patch fixes it.
Signed-off-by: Wengang Wang
You will have to trace thru process_blocked_lock() to make sense of this.
Coly Li wrote:
Sunil Mushran Wrote:
So the thread is removing the lockres from the list (thus making
the count 0), and then calling ocfs2_process_blocked_lock() which
is adding that lockres back in the list (thus 1
Joel Becker wrote:
ocfs2 has a problem with open(O_CREAT|O_EXCL). Once you've created the
file, you can't restart the open(), because O_CREAT|O_EXCL will trigger
-EEXIST.
The problem is that ocfs2 is catching the signal -permission(), called
by may_open(). This happens after -create() has
How about we narrow down the issue by dumping the lockres?
Look in 1.4. We dump the lockres in dlm_purge_lockres().
__dlm_print_one_lock_resource(res);
In this case, it appears the user has encountered it more than
once. Work with Srini to give them a package with the above.
The idea here is to
It's too intrusive for a debug patch. And big too. Maybe we need this
eventually. But for right now, it will be easier if we can narrow down
the scope of the issue as quickly as possible.
On Sep 2, 2009, at 8:21 PM, Wengang Wang wen.gang.w...@oracle.com
wrote:
Hi Sunil,
Sunil Mushran
Signed-off-by: Sunil Mushran sunil.mush...@oracle.com
Tao Ma wrote:
In commit a5a0a630922a2f6a774b6dac19f70cb5abd86bb0, when
ocfs2_attch_dentry_lock fails, we call an extra iput and reset
dentry-d_fsdata to NULL. This resolve a bug, but it isn't
completed and the dentry is still there. When
Signed-off-by: Sunil Mushran sunil.mush...@oracle.com
Goldwyn Rodrigues wrote:
In case a downconvert is queued, and a flock receives a signal,
BUG_ON(lockres-l_action != OCFS2_AST_INVALID) is triggered
because a lock cancel triggers a dlmunlock while an AST is
scheduled.
To avoid
defers the handling to the mount option, errors.
Signed-off-by: Sunil Mushran sunil.mush...@oracle.com
---
fs/ocfs2/super.c |3 ++-
1 files changed, 2 insertions(+), 1 deletions(-)
diff --git a/fs/ocfs2/super.c b/fs/ocfs2/super.c
index b0ee0fd..44a0388 100644
--- a/fs/ocfs2/super.c
+++ b/fs/ocfs2
1.4.0 is very old. sles needs to refresh the tools.
Peter W. Morreale wrote:
Hi all, I had the following happen when playing with adding the -C
option to mkfs.ocfs2 commandline. Is this a known issue? Should I
file a bug report?
Works fine as long as I do not specify a cluster size.
In a non-sparse extend, we correctly allocate (and zero) the clusters between
the old_i_size and pos, but we don't zero the portions of the cluster we're
writing to outside of pos-len.
It handles clustersize pagesize and blocksize pagesize.
[Cleaned up by Joel Becker.]
Signed-off-by: Sunil
So the problem with the last patch was that it did not handle blocksize
pagesize.
On a x86 box, with a 2K/4K file system, it was only initializing the first 2K
block.
This fixes it.
Sunil
___
Ocfs2-devel mailing list
Ocfs2-devel@oss.oracle.com
In a non-sparse extend, we correctly allocate (and zero) the clusters between
the old_i_size and pos, but we don't zero the portions of the cluster we're
writing to outside of pos-len.
[Cleaned up by Joel Becker.]
Signed-off-by: Sunil Mushran sunil.mush...@oracle.com
---
fs/ocfs2/aops.c | 54
The patch was created against a 1.4 tree. However, it applies cleanly to
mainline too.
The patch has been lightly tested. I am running fill_verify_holes on a non
sparse volume
currently.
Please review.
Sunil
___
Ocfs2-devel mailing list
Can you explain why you think it is redundant?
Goldwyn Rodrigues wrote:
Remove redundant BUG_ON()
Signed-off-by: Goldwyn Rodrigues rgold...@suse.de
---
diff --git a/fs/ocfs2/dlm/dlmast.c b/fs/ocfs2/dlm/dlmast.c
index d07ddbe..81eff8e 100644
--- a/fs/ocfs2/dlm/dlmast.c
+++
Signed-off-by: Sunil Mushran sunil.mush...@oracle.com
Goldwyn Rodrigues wrote:
Remove redundant BUG_ON()
Signed-off-by: Goldwyn Rodrigues rgold...@suse.de
---
diff --git a/fs/ocfs2/dlm/dlmast.c b/fs/ocfs2/dlm/dlmast.c
index d07ddbe..81eff8e 100644
--- a/fs/ocfs2/dlm/dlmast.c
+++ b
Wengang,
Idea is to print stuff that will be useful. Instead of printing
all args, you can always prepend a 3 char code... flg=0x%x to
identify the args that you are printing. Remember the printk
buffer is limited.
Thanks for taking on this task.
Sunil
Wengang Wang wrote:
Hi Joel,
Joel
Thanks Jan. It has been pushed to the quotas-support branch.
We'll start the review/testing process.
Jan Kara wrote:
Hello Sunil,
attached is a patch which implements full quota support for ocfs2 tools.
It's based against ocfs2-tools git's master branch. Would you please
consider
Goldwyn Rodrigues wrote:
generic_write_checks() expects count to be initialized to the size of
the write.
Writes to files open with O_DIRECT|O_LARGEFILE write 0 bytes because
count in uninitialized.
Signed-off-by: Goldwyn Rodrigues rgold...@suse.de
Signed-off-by: Sunil Mushran
Tao Ma wrote:
Hi wengang,
Wengang Wang wrote:
Sunil,
in the src, I see both mlog(ML_ERROR...)(finally printk(KERN_ERR...))
and ocfs2_error()(finally printk(KERN_CRIT...).
could you tell me in what case one should be used?
mlog(ML_ERROR...) only print errors. But ocfs2_error does other
sob
Hi Wengang, functions is aops.c are missing mlogs. Do you have the
bandwidth to help out?
On Jul 9, 2009, at 10:26 PM, Wengang Wang wen.gang.w...@oracle.com
wrote:
in ocfs2_file_aio_write(), log_exit() could don't log the value
which is really returned. this patch fixes it.
David Teigland wrote:
Yes, the kernel dlm api allows a variable lvb size, but the user dlm api fixes
it at 32.
Do you need to actually use a 64 byte lvb from userspace? Or do you just need
to create the locksapce with a 64 byte lvb? We could add a flag to work
around the later fairly
Signed-off-by: Sunil Mushran sunil.mush...@oracle.com
Jeff Liu wrote:
in dlmrecovery.c:1121, replace 'migrate' to 'migration' to keep the
consistency
by comparing to other lines with the similar log info in the same file.
Signed-off-by: Jeff Liu jeff@oracle.com
---
fs/ocfs2/dlm
Jeff Mahoney wrote:
There's a problem with this patch that I ran into while testing
2.6.31-rc2. ocfs2_orphan_scan_stop expects that ocfs2_orphan_scan_init
was called. If the mount fails for any reason, ocfs2_dismount_volume
calls ocfs2_orphan_scan_stop. The thing is that ocfs2_dismount_volume
Jeff Mahoney wrote:
Yup, you're right. Fixed version attached.
Signed-off-by: Sunil Mushran sunil.mush...@oracle.com
___
Ocfs2-devel mailing list
Ocfs2-devel@oss.oracle.com
http://oss.oracle.com/mailman/listinfo/ocfs2-devel
Tools is enough. It supports all existing versions of the kernel fs.
On Jul 4, 2009, at 6:10 AM, syla...@aim.com wrote:
Hi,
Thanks :)
BTW, Do I need to upgrade the kernel module too, or just the ocfs2-
tools package?
If I need to upgrade the module too ... I am having trouble
compiling
TaoMa wrote:
I've read fs/ocfs2/dlm/dlmcommon.h to study the structure of
dlm_migratable_lockres.
However, I may find a typo for the DLM_MIG_LOCKRES_MAX_LEN,
the NET_MAX_PAYLOAD_BYTES should be O2NET_MAX_PAYLOAD_BYTES, I think.
In comments, the sizeof(net_msg) should be sizeof(o2net_msg) by
501 - 600 of 958 matches
Mail list logo
