Good catch! Thank you for the fix
Reviewed-by: Srinivas Eeda <srinivas.e...@oracle.com>
On 08/11/2016 04:12 PM, Ashish Samant wrote:
> If we do fallocate with punch hole option on a reflink, with start offset
> on a cluster boundary and end offset somewhere in another cluster, w
In simple terms, mle's get into life at the beginning of the lock
mastery and ends at the end of lock mastery(either new lockres or during
lockres migration). It's purpose is to handle the race when more than
one node try to master or find master of a lockres.
On 05/03/2016 02:23 AM,
On 03/28/2016 05:41 AM, jiangyiwen wrote:
> On 2016/3/28 9:23, Junxiao Bi wrote:
>> Hi Yiwen,
>>
>> On 03/26/2016 10:54 AM, jiangyiwen wrote:
>>> Hi, Junxiao
>>> This patch may have a problem. That is journal of every nodes become
>>> abort when storage down, and then when storage up, because
Thanks Junxiao!
Acked-by: Srinivas Eeda <srinivas.e...@oracle.com>
On 12/06/2015 08:09 PM, Junxiao Bi wrote:
> commit 8f1eb48758aa ("ocfs2: fix umask ignored issue") introduced an issue,
> SGID of sub dir was not inherited from its parents dir. It is because SGID
> i
On 11/24/2015 01:46 PM, Mark Fasheh wrote:
> On Tue, Nov 03, 2015 at 04:20:27PM +0800, Junxiao Bi wrote:
>> Hi Gang,
>>
>> On 11/03/2015 03:54 PM, Gang He wrote:
>>> Hi Junxiao,
>>>
>>> Thank for your reviewing.
>>> Current design, we use a sysfile as a interface to check/fix a file (via
>>> pass
Hi Gang,
thanks for pointing to explanation of the feature.
What I am curious about is ... what were the real cases that you came
across prompted this change and how this change would help in that case.
Thanks,
--Srini
On 10/28/2015 09:44 PM, Gang He wrote:
> Hello Srini,
>
> There is a doc
Hi Gang,
thank you for implementing this. I would like to understand this better
on where and how it helps ... would you mind sharing couple
examples(real scenarios).
Thanks,
--Srini
On 10/27/2015 11:25 PM, Gang He wrote:
> When there are errors in the ocfs2 filesystem,
> they are usually
Thank you all, for all you contributions in fixing ocfs2 test suite,
nice effort :)
On 10/08/2015 11:09 PM, Junxiao Bi wrote:
> Hi Eric,
>
> Thank you for reviewing patches and make a summary of the project. Let
> put more effort to make it better in the future.
>
> Thanks,
> Junxiao.
>
> On
Hi Goldwyn
On 09/01/2015 03:32 AM, Goldwyn Rodrigues wrote:
> Hi Junxiao,
>
> On 08/31/2015 09:22 PM, Junxiao Bi wrote:
>> Hi Goldwyn,
>>
>> Ocfs2 kernel version is removed from commit
>> ff8fb335221e2c446b0d4cbea26be371fd2feb64 ("ocfs2: remove versioning
>> information"), but Oracle CRS depends
Hi Norton,
while the localalloc bmp is enabled the chances of a particular gd
ending with zero free bits is very minimal. local alloc bmp will pick
next gd once min number of free bits falls below localalloc bmp size. So
next gd is picked while the current gd still has free space.
Having said
On 08/20/2015 04:50 AM, Norton.Zhu wrote:
Currently error handling in dlm_request_join is a little obscure.
So optimize it to promote readability.
Signed-off-by: Norton.Zhu norton@huawei.com
---
fs/ocfs2/dlm/dlmdomain.c | 69
++--
1 file
Ok, understood. Thanks for explaining :)
On 01/21/2015 05:07 PM, Joseph Qi wrote:
Hi Srini,
On 2015/1/22 1:55, Srinivas Eeda wrote:
Hi Joesph,
thanks a lot for submitting the above patch. I am trying to understand what
kind of flag corruption have you noticed and under what circumstances
git repos for ocfs2 kernel code is linux.git and linux-next.git on
kernel.org
For tools it's git://oss.oracle.com/git/ocfs2-tools.git
On 01/08/2015 10:43 AM, Jonathan Fraser wrote:
Hi,
Sorry for the newbe question, but where are the git repos for the staged
changes? I've searched high and
Hi Goldwyn, Germano
I have updated ocfs2-tools git repo with latest fixes :)
Germano,
following documentation is available for OCFS2
http://docs.oracle.com/cd/E37670_01/E37355/html/ol_ocfs2.html
http://www.oracle.com/us/technologies/linux/025995.htm
looks good. Thanks for your explanation and fix
Reviewed-by: Srinivas Eeda srinivas.e...@oracle.com
On 10/30/2014 11:08 PM, Junxiao Bi wrote:
Set nn_persistent_error to -ENOTCONN will stop reconnect since the
stop condition in o2net_start_connect() will be true.
stop = (nn-nn_sc
,
--Srini
On 10/29/2014 10:32 PM, Junxiao Bi wrote:
Hi Srini,
On 10/30/2014 01:16 PM, Srinivas Eeda wrote:
Junxiao,
can you please describe under what circumstances you saw this
problem?
My understanding is o2net_connect_expired is only queued when
connection
actually broke and ENOTCONN
Junxiao,
can you please describe under what circumstances you saw this problem?
My understanding is o2net_connect_expired is only queued when connection
actually broke and ENOTCONN is the right error in that case.
Thanks,
--Srini
On 10/29/2014 06:41 PM, Junxiao Bi wrote:
Set
Hi Junxiao,
thanks for explaining. For this case allowing a reconnect (setting
atomic_set(nn-nn_timeout, 1); ) in o2net_connect_expired should work ?
Thanks,
--Srini
On 10/29/2014 10:32 PM, Junxiao Bi wrote:
Hi Srini,
On 10/30/2014 01:16 PM, Srinivas Eeda wrote:
Junxiao,
can you please
will respond they are not the master. Node C then masters
the resource and sends assert master to all nodes. This will now make node A
crash with the following message.
dlm_assert_master_handler:1831 ERROR: DIE! Mastery assert from 9, but current
owner is 10!
Signed-off-by: Srinivas Eeda srinivas.e
This fix will cause problems for regular file unlinks. If the inode is
not cleared on node 2, then both node 1 and node 2 will fail to get try
open lock and fail to clear the orphan inodes.
I am assuming the deadlock you have seen is because of quota's enabled
and fix
On 10/09/2014 12:53 AM, Joseph Qi wrote:
On 2014/10/9 15:16, Guozhonghua wrote:
Hi Joseph and Srinivas,
We had merged and test the two patches:
1.ocfs2: o2net: set tcp user timeout to max value
8e9801dfe37c9e68cdbfcd15988df2187191864e
2.ocfs2: o2net: don't shutdown
Joseph,
yes I agree with you that the proposed change won't work for all cases.
The only way to fix if we allow reconnects is to track all messages and
replay the last few in case of reconnects.
However this problem shouldn't arise because of the following two fixes
Junxiao put in. Can you
Looks good to me. Thanks for the patch
Reviewed-by: Srinivas Eeda srinivas.e...@oracle.com
On 09/15/2014 10:15 PM, Junxiao Bi wrote:
Firing quorum before connection established can cause unexpected node to
reboot.
Assume there are 3 nodes in the cluster, Node 1, 2, 3. Node 2 and 3 have
Acked-by: Srinivas Eeda srinivas.e...@oracle.com
On 06/01/2014 06:53 AM, Rickard Strandqvist wrote:
There is a risk that the variable will be used without being initialized.
This was largely found by using a static code analysis program called
cppcheck.
Signed-off-by: Rickard Strandqvist
Hi Andrew,
can you please pull the following patch from Tetsuo Handa. It fixes a
regression in ocfs2/mainline and linux-next
Thanks,
--Srini
___
Ocfs2-devel mailing list
Ocfs2-devel@oss.oracle.com
https://oss.oracle.com/mailman/listinfo/ocfs2-devel
kfree(xattr-name) from ocfs2_mknod() and ocfs2_symlink().
Reported-by: Tariq Saeed tariq.x.sa...@oracle.com
Signed-off-by: Tetsuo Handa penguin-ker...@i-love.sakura.ne.jp
Reviewed-by: Srinivas Eeda srinivas.e...@oracle.com
---
fs/ocfs2/namei.c |2 --
1 file changed, 2 deletions(-)
diff --git
Thanks for explaining
Reviewed-by: Srinivas Eeda srinivas.e...@oracle.com
On 03/25/2014 06:25 PM, alex chen wrote:
On 2014/3/26 2:45, Srinivas Eeda wrote:
These changes looks good to me. However ocfs2_read_blocks and
ocfs2_read_blocks_sync needs the same fix ? :)
There is no need to do
These changes looks good to me. However ocfs2_read_blocks and
ocfs2_read_blocks_sync needs the same fix ? :)
On 03/25/2014 12:05 AM, alex chen wrote:
Do not put bh when buffer_uptodate failed in ocfs2_write_block and
ocfs2_write_super_or_backup, because it will put bh in b_end_io.
Otherwise
Guozhonghua, idea of the fix looks good, but the patch doesn't seem to
meet mainline standards. Can you please follow the instructions that
Jeff emailed you last time and resubmit the patch.
On 03/11/2014 04:09 AM, Guozhonghua wrote:
I review the code of the kernel 3.11.10.
The difference
...@gmail.com
Cc: Srinivas Eeda srinivas.e...@oracle.com
Cc: sta...@vger.kernel.org
---
fs/ocfs2/dlm/dlmrecovery.c | 14 --
1 file changed, 8 insertions(+), 6 deletions(-)
diff --git a/fs/ocfs2/dlm/dlmrecovery.c b/fs/ocfs2/dlm/dlmrecovery.c
index 7035af0..c2dd258 100644
--- a/fs
Junxiao, thanks for looking into this issue. Please see my comment below
On 02/24/2014 01:07 AM, Junxiao Bi wrote:
Hi,
On 07/19/2012 09:59 AM, Sunil Mushran wrote:
Different issues.
On Wed, Jul 18, 2012 at 6:34 PM, Junxiao Bi junxiao...@oracle.com
mailto:junxiao...@oracle.com wrote:
Hi Jan,
thanks a lot for these patches. They all look good to me ... I just have
one question on patch 5
Thanks,
--Srini
On 02/20/2014 07:18 AM, Jan Kara wrote:
Hello,
here is a second version of my patchset to solve a deadlocks when we do not
defer dropping of inode reference from
looks good to me
Reviewed-by: Srinivas Eeda srinivas.e...@oracle.com
On 02/20/2014 07:18 AM, Jan Kara wrote:
Provide dqgrab() function to get quota structure reference when we are
sure it already has at least one active reference. Make use of this
function inside quota code.
Signed-off
looks good to me
Reviewed-by: Srinivas Eeda srinivas.e...@oracle.com
On 02/20/2014 07:18 AM, Jan Kara wrote:
We cannot drop last dquot reference from downconvert thread as that
creates the following deadlock:
NODE 1 NODE2
holds dentry lock for 'foo'
holds
Reviewed-by: Srinivas Eeda srinivas.e...@oracle.com
On 02/20/2014 07:18 AM, Jan Kara wrote:
The flag was never set, delete it.
Signed-off-by: Jan Kara j...@suse.cz
---
fs/ocfs2/inode.c | 6 --
fs/ocfs2/inode.h | 8 +++-
fs/ocfs2/journal.c | 6 --
3 files changed, 3
looks good to me
Reviewed-by: Srinivas Eeda srinivas.e...@oracle.com
On 02/20/2014 07:18 AM, Jan Kara wrote:
Move dquot_initalize() call in ocfs2_delete_inode() after the moment we
verify inode is actually a sane one to delete. We certainly don't want
to initialize quota for system inodes etc
I like the idea of dc_task handling queued basts in
ocfs2_mark_lockres_freeing.
I am wondering if we should call lockres-l_ops-post_unlock(osb,
lockres) ? Would there be another node waiting for a bast response ?
On 02/20/2014 07:18 AM, Jan Kara wrote:
If we are dropping last inode reference
On 01/24/2014 01:55 PM, Mark Fasheh wrote:
On Fri, Jan 24, 2014 at 12:47:02PM -0800, a...@linux-foundation.org wrote:
From: Tariq Saeed tariq.x.sa...@oracle.com
Subject: ocfs2/o2net: incorrect to terminate accepting connections loop upon
rejecting an invalid one
When o2net-accept-one()
Hi Goldwyn,
I am not 100% sure ... but my understanding is ocfs2_fs.h defines
metadata structures and helper functions. This file is shared between
ocfs2-tools and kernel modules. The functions you mentioned are used by
ocfs2-tools code and hence exist in this file.
Thanks,
--Srini
On
On 01/20/2014 07:31 AM, Goldwyn Rodrigues wrote:
On 01/16/2014 04:58 PM, Jan Kara wrote:
On Thu 16-01-14 23:28:49, Jan Kara wrote:
We cannot drop last dquot reference from downconvert thread as that
creates the following deadlock:
NODE 1 NODE2
holds dentry
Hi Jan,
thanks a lot for explaining the problem. Please see my comment below.
On 01/16/2014 06:02 AM, Jan Kara wrote:
On Thu 16-01-14 07:35:58, Goldwyn Rodrigues wrote:
On 01/15/2014 08:47 PM, Jan Kara wrote:
On Wed 15-01-14 17:17:55, Goldwyn Rodrigues wrote:
On 01/15/2014 09:53 AM, Jan Kara
On 01/13/2014 08:06 PM, Joseph Qi wrote:
On 2014/1/11 9:19, Srinivas Eeda wrote:
From: Srinivas Eeda seeda@srini.(none)
A tiny race between BAST and unlock message causes the NULL dereference.
A node sends an unlock request to master and receives a response. Before
processing the response
On 01/13/2014 07:37 AM, Joel Becker wrote:
On Fri, Jan 10, 2014 at 05:19:13PM -0800, Srinivas Eeda wrote:
From: Srinivas Eeda seeda@srini.(none)
A tiny race between BAST and unlock message causes the NULL dereference.
A node sends an unlock request to master and receives a response. Before
From: Srinivas Eeda seeda@srini.(none)
A tiny race between BAST and unlock message causes the NULL dereference.
A node sends an unlock request to master and receives a response. Before
processing the response it receives a BAST from the master. Since both requests
are processed by different
On 01/09/2014 08:34 AM, Goldwyn Rodrigues wrote:
On 01/09/2014 10:06 AM, Srinivas Eeda wrote:
On 01/09/2014 07:44 AM, Goldwyn Rodrigues wrote:
Hi Srini,
Thanks for the reply.
On 01/08/2014 11:30 PM, Srinivas Eeda wrote:
From the comments in fs/ocfs2/inode.h:90 it seems, this was
used
Hi Goldwyn,
On 01/08/2014 04:12 PM, Goldwyn Rodrigues wrote:
Hi,
From the comments in fs/ocfs2/inode.h:90 it seems, this was used in
legacy ocfs2 systems when a node received unlink votes. Since unlink
votes has been done away with and replaced with open locks, is this
flag still required?
On 01/08/2014 07:12 PM, Goldwyn Rodrigues wrote:
Hi Srini,
On 01/08/2014 07:29 PM, Srinivas Eeda wrote:
Hi Goldwyn,
On 01/08/2014 04:12 PM, Goldwyn Rodrigues wrote:
Hi,
From the comments in fs/ocfs2/inode.h:90 it seems, this was used in
legacy ocfs2 systems when a node received unlink
is unset by migrate handler thread. umount
thread will ultimately migrate this lockres during another pass of the
lockres hash list.
Signed-off-by: Tariq Saeed tariq.x.sa...@oracle.com
Signed-off-by: Srinivas Eeda srinivas.e...@oracle.com
---
fs/ocfs2/dlm/dlmmaster.c |4
1 files
On 05/31/2013 03:38 AM, Xue jiufei wrote:
Hi, Xiaowei
It's OK to simlify the patch just as you did. But we don't want to resend
master request to all others nodes in consideration of network traffic.
So we record those maybe down nodes in down_nodemap.
δΊ 2013/5/28 14:12, xiaowei.hu ει:
Hi,
Hi Joel, et al,
sorry for the delay in resending discontiguous localalloc patches. Can you
please review when you get a chance. I'll email the tools patches once kernel
patches are approved.
I came across two use cases where this feature will help.
1. On a customer site, after running an
will try to look for
atleast that big of chunk. If it's not available then it reduces the size by
half and retries.
#define OCFS2_DEFAULT_LOCALALLOC_CHUNK_SIZE (128*1024) /* 128kb */
Signed-off-by: Srinivas Eeda srinivas.e...@oracle.com
---
fs/ocfs2/localalloc.c | 28
of m_bitmap_ext_arr
Callers should pass number of discontiguous chunks during ocfs2_resmap_restart
and later call ocfs2_resmap_set_extent_size for every chunk.
Signed-off-by: Srinivas Eeda srinivas.e...@oracle.com
---
fs/ocfs2/localalloc.c |4 +++-
fs/ocfs2/reservations.c | 39
and as the filesystem gets
fragmented you may see multipe records.
This feature can be enabled/disabled when the file system is offline.
Signed-off-by: Srinivas Eeda srinivas.e...@oracle.com
---
fs/ocfs2/localalloc.c | 22 ++--
fs/ocfs2/ocfs2.h |7 +
fs/ocfs2/ocfs2_fs.h
This patch adds code to support discontiguous localalloc bitmap. At any given
time there can be a combination of volumes that have discontigous feature
enabled or disabled.
Signed-off-by: Srinivas Eeda srinivas.e...@oracle.com
---
fs/ocfs2/localalloc.c | 478
ocfs2_local_alloc_inode is used in multiple functions. It is convenient if we
move ocfs2_local_alloc_inode to ocfs2 super.
Signed-off-by: Srinivas Eeda srinivas.e...@oracle.com
---
fs/ocfs2/localalloc.c | 53 -
fs/ocfs2/ocfs2.h |1 +
2
On 8/22/2012 7:17 AM, Jie Liu wrote:
Hi All,
These days, I am investigating an issue regarding OCFS2 unexpected
reboot in some real world use cases.
This problem occurred when the network status goes south, when the
disk IO load is too high, etc...
I suspect it might caused by ocfs2 fencing
Hi Jan,
thanks for helping.
Jan Kara wrote:
Hello,
his comments:
@ With those patches in, all other nodes will now queue downgrade of dentry
@ locks to ocfs2_wq thread. Then Node 1 gets a lock is in use when it calls
@ ocfs2_try_open_lock and so does other nodes and hence orphans lie
comments inline
On 5/24/2012 10:53 PM, xiaowei...@oracle.com wrote:
From: Xiaowei.Huxiaowei...@oracle.com
when the master requested locks ,but one/some of the live nodes died,
after it received the request msg and before send out the locks packages,
the recovery will fall into endless
of
m_bitmap_ext_arr.
Signed-off-by: Srinivas Eeda srinivas.e...@oracle.com
---
fs/ocfs2/reservations.c | 41 ++---
fs/ocfs2/reservations.h |7 ++-
2 files changed, 40 insertions(+), 8 deletions(-)
diff --git a/fs/ocfs2/reservations.c b/fs/ocfs2
chunks. In
best case there is only one record and increases as the filesystem gets
fragmented. Number of records at a time are limited depending on the size
of the bitmap and the max limit is defined by OCFS2_MAX_LOCAL_ALLOC_RECS.
Signed-off-by: Srinivas Eeda srinivas.e...@oracle.com
---
fs/ocfs2
Hi all,
can you please review following 3 patches that implement discontiguous
localalloc bitmap support for ocfs2 file system. This feature helps
applications that significantly fragment the filesystem.
These fixes needs changes to ocfs2 tools as well. I am sending those patches
for review
This patch adds supporting functions and modifies localalloc code to implement
discontiguous localalloc bitmap.
Signed-off-by: Srinivas Eeda srinivas.e...@oracle.com
---
fs/ocfs2/localalloc.c | 523 -
1 files changed, 342 insertions(+), 181
Joel Becker wrote:
On Mon, May 07, 2012 at 04:21:27PM -0700, Srinivas Eeda wrote:
can you please review following 3 patches that implement discontiguous
localalloc bitmap support for ocfs2 file system. This feature helps
applications that significantly fragment the filesystem
Joel Becker wrote:
On Mon, May 07, 2012 at 04:21:29PM -0700, Srinivas Eeda wrote:
OH MY DOG NO. NEVER EVER DO THIS. You cannot update an old
filesystem on the fly! What about other nodes that are running older
versions of the software? They will crash or corrupt data
]
Signed-off-by: Srinivas Eeda srinivas.e...@oracle.com
---
fs/ocfs2/dlmglue.c | 30 ++
1 files changed, 18 insertions(+), 12 deletions(-)
diff --git a/fs/ocfs2/dlmglue.c b/fs/ocfs2/dlmglue.c
index 81a4cd2..d8552a5 100644
--- a/fs/ocfs2/dlmglue.c
+++ b/fs/ocfs2
]
Signed-off-by: Srinivas Eeda srinivas.e...@oracle.com
---
fs/ocfs2/dlmglue.c | 31 +++
1 files changed, 19 insertions(+), 12 deletions(-)
diff --git a/fs/ocfs2/dlmglue.c b/fs/ocfs2/dlmglue.c
index 81a4cd2..67af5db 100644
--- a/fs/ocfs2/dlmglue.c
+++ b/fs/ocfs2
sorry ignore this patch, resent another one after adding the new line.
On 1/30/2012 9:47 PM, Srinivas Eeda wrote:
When ocfs2dc thread holds dc_task_lock spinlock and receives soft IRQ it
deadlock itself trying to get same spinlock in ocfs2_wake_downconvert_thread.
Below is the stack snippet
[ocfs2_nodemanager]
[a030aac8] o2net_rx_until_empty+0x762/0x90d [ocfs2_nodemanager]
[81071802] worker_thread+0x14d/0x1ed
Signed-off-by: Srinivas Eeda srinivas.e...@oracle.com
---
fs/ocfs2/dlm/dlmast.c |3 ++-
1 files changed, 2 insertions(+), 1 deletions(-)
diff --git a/fs/ocfs2/dlm
Hi Tao,
thanks for reviewing.
When ocfs2dc thread holds dc_task_lock spinlock and receives soft IRQ for
I/O completion it deadlock itself trying to get same spinlock in
ocfs2_wake_downconvert_thread
could you please describe it in more detail?
When ocfs2dc thread is running on a cpu and
When ocfs2dc thread holds dc_task_lock spinlock and receives soft IRQ for
I/O completion it deadlock itself trying to get same spinlock in
ocfs2_wake_downconvert_thread
The patch disables interrupts when acquiring dc_task_lock spinlock
Signed-off-by: Srinivas Eeda srinivas.e...@oracle.com
node 2's orphan dir. This
patch exists because that wasn't happening.
On 7/7/2011 1:26 PM, Sunil Mushran wrote:
On 07/07/2011 01:02 PM, Sunil Mushran wrote:
On 07/06/2011 11:19 PM, Srinivas Eeda wrote:
On 7/5/2011 11:17 PM, Sunil Mushran wrote:
2. All nodes have to scan all slots. Even live
On 7/5/2011 11:17 PM, Sunil Mushran wrote:
2. All nodes have to scan all slots. Even live slots. I remember we
did for
a reason. And that reason should be in the comment in the patch written
by Srini.
When a node unlinks a file it inserts an entry into it's own orphan
slot. If another node
+ spin_unlock(tmpres-spinlock);
+ spin_unlock(dlm-spinlock);
lockres could still get added to purgelist at this point and we
could still have the same problem? I think, here we need some
mechanism that marks the lockres is in use that would protect it
from adding to the
I think I have seen this problem in ocfs2-1.2 and it was addressed by
using a new state DLM_LOCK_RES_IN_USE. But we didn't merge into mainline
as sunil suggested we need to look for a different approach
http://oss.oracle.com/pipermail/ocfs2-devel/2010-June/006669.html
From: Sunil Mushran sunil.mush...@oracle.com
Mainline fa16655a622e7c0fda76ca5155db6efc86968c65
A per-region debugfs file, elapsed_time_in_ms, shows the time since the
heartbeat timer was last armed.
Signed-off-by: Sunil Mushran sunil.mush...@oracle.com
---
fs/ocfs2/cluster/heartbeat.c | 46
The following two patches are backports from mainline to 1.4. These patches
create debugfs entry for heartbeat regions and to show elapsed time.
___
Ocfs2-devel mailing list
Ocfs2-devel@oss.oracle.com
http://oss.oracle.com/mailman/listinfo/ocfs2-devel
From: Sunil Mushran sunil.mush...@oracle.com
Mainline 0841ed580fe8a3e51ba9dbb133dafc787cce428f
Signed-off-by: Sunil Mushran sunil.mush...@oracle.com
---
fs/ocfs2/cluster/heartbeat.c | 27 +++
1 files changed, 19 insertions(+), 8 deletions(-)
diff --git
## Block#TotalUsed Free Contig Size
052803072 322564294965350 34202182074032
..
Signed-off-by: Srinivas Eeda srinivas.e...@oracle.com
---
fs/ocfs2/suballoc.c | 16
1 files changed, 16 insertions(+), 0
## Block#TotalUsed Free Contig Size
052803072 322564294965350 34202182074032
..
Signed-off-by: Srinivas Eeda srinivas.e...@oracle.com
---
fs/ocfs2/suballoc.c | 16
1 files changed, 16 insertions(+), 0
as leaving. Only block mles are expected
to be around, so we log ERROR for other mles but still free them.
Signed-off-by: Srinivas Eeda srinivas.e...@oracle.com
---
fs/ocfs2/dlm/dlmcommon.h |1 +
fs/ocfs2/dlm/dlmdomain.c |1 +
fs/ocfs2/dlm/dlmmaster.c | 34 ++
3
as leaving. Only block mles are expected
to be around, so we log ERROR for other mles but still free them.
Signed-off-by: Srinivas Eeda srinivas.e...@oracle.com
---
fs/ocfs2/dlm/dlmcommon.h |1 +
fs/ocfs2/dlm/dlmdomain.c |1 +
fs/ocfs2/dlm/dlmmaster.c | 40
on master list after the node sent
DLM_EXIT_DOMAIN_MSG to other nodes.
Signed-off-by: Srinivas Eeda srinivas.e...@oracle.com
---
fs/ocfs2/dlm/dlmcommon.h |1 +
fs/ocfs2/dlm/dlmdomain.c |1 +
fs/ocfs2/dlm/dlmmaster.c | 33 +
3 files changed, 35 insertions
thanks for making this patch, it looks good just few minor changes about
comments
On 7/23/2010 5:15 AM, Wengang Wang wrote:
In the following situation, there remains an incorrect bit in refmap on the
recovery master. Finally the recovery master will fail at purging the lockres
due to the
protecting it from getting reused.
Signed-off-by: Srinivas Eeda srinivas.e...@oracle.com
Acked-by: Sunil Mushran sunil.mush...@oracle.com
---
fs/ocfs2/dlm/dlmthread.c | 80 +++--
1 files changed, 34 insertions(+), 46 deletions(-)
diff --git a/fs/ocfs2/dlm
Joel Becker wrote:
On Tue, Jun 22, 2010 at 10:48:28PM -0700, Srinivas Eeda wrote:
+if (!__dlm_lockres_unused) {
+mlog(ML_ERROR, found lockres %s:%.*s: in use after deref\n,
+ dlm-name, res-lockname.len, res-lockname.name
it from getting reused.
Signed-off-by: Srinivas Eeda srinivas.e...@oracle.com
Signed-off-by: Sunil Mushransunil.mush...@oracle.com
---
fs/ocfs2/dlm/dlmthread.c | 79 +++--
1 files changed, 33 insertions(+), 46 deletions(-)
diff --git a/fs/ocfs2/dlm
Sunil, Joel, Wengang. Thanks for reviewing the patch and your comments.
On 6/23/2010 10:00 AM, Sunil Mushran wrote:
Signed-off-by: Sunil Mushransunil.mush...@oracle.com
On 06/22/2010 10:48 PM, Srinivas Eeda wrote:
There are two problems in dlm_run_purgelist
1. If a lockres is found
it from getting reused.
Signed-off-by: Srinivas Eeda srinivas.e...@oracle.com
---
fs/ocfs2/dlm/dlmthread.c | 76 --
1 files changed, 33 insertions(+), 43 deletions(-)
diff --git a/fs/ocfs2/dlm/dlmthread.c b/fs/ocfs2/dlm/dlmthread.c
index 11a6d1f..cb74689
it from getting reused.
Signed-off-by: Srinivas Eeda srinivas.e...@oracle.com
---
fs/ocfs2/dlm/dlmthread.c | 79 +++--
1 files changed, 33 insertions(+), 46 deletions(-)
diff --git a/fs/ocfs2/dlm/dlmthread.c b/fs/ocfs2/dlm/dlmthread.c
index 11a6d1f..6822f9a
it from getting reused.
Signed-off-by: Srinivas Eeda srinivas.e...@oracle.com
---
fs/ocfs2/dlm/dlmthread.c | 55 +
1 files changed, 26 insertions(+), 29 deletions(-)
diff --git a/fs/ocfs2/dlm/dlmthread.c b/fs/ocfs2/dlm/dlmthread.c
index 11a6d1f..79d1ef6
is adding
more messiness to an area that is already very messy.
Sunil
On 06/15/2010 09:43 PM, Srinivas Eeda wrote:
This patch fixes the following hole.
dlmlock tries to create a new lock on a lockres that is on purge
list. It calls
dlm_get_lockresource and later adds a lock to blocked list
On 6/17/2010 1:32 AM, Joel Becker wrote:
On Wed, Jun 16, 2010 at 06:44:43PM -0700, Sunil Mushran wrote:
One way to skip a lockres in the purgelist is to list_del_init() and
list_add_tail(). That simplifies the patch a lot.
I have attached a quick dirty patch. See if that satisfies all the
On 6/17/2010 7:48 AM, Sunil Mushran wrote:
On 06/17/2010 01:35 AM, Srinivas Eeda wrote:
On 6/17/2010 1:32 AM, Joel Becker wrote:
On Wed, Jun 16, 2010 at 06:44:43PM -0700, Sunil Mushran wrote:
One way to skip a lockres in the purgelist is to list_del_init() and
list_add_tail
Sunil,
as of now, there is still a window in dlm_get_lock_resource, where it
finds the lockres but it doesn't protect it from getting purged. Second
patch fixes this by marking it in_use, can you please review that one as
well.
Thanks,
--Srini
On 6/17/2010 8:06 AM, Sunil Mushran wrote:
On
patch looks good, it fixes the umount code path which prevents a lockres
from migrating if it needs to be recovered. I have few comments on the
scenario you described.
On 6/11/2010 3:25 AM, Wengang Wang wrote:
Any comment on this patch?
regards,
wengang.
On 10-05-25 15:59, Wengang Wang
from the master lockres is not found
This patch marks the lockres with a new state DLM_LOCK_RES_IN_USE which would
protect lockres from dlm_thread purging it.
Signed-off-by: Srinivas Eeda srinivas.e...@oracle.com
---
fs/ocfs2/dlm/dlmcommon.h |1 +
fs/ocfs2/dlm/dlmlock.c |4
fs/ocfs2
The lock order in this code causes dead lock, not caused by your patch.
The lock order in dlm_query_join_handler is
dlm_domain_lock
-dlm-spinlock
dead locks with ..
dlm_lockres_put calls dlm_lockres_release while holding dlm-spinlock
which calls dlm_put which gets dlm_domain_lock. So the
Please ignore this patch. I'll resend this patch along with dlm purge
lockres for completeness.
On 6/9/2010 7:07 PM, Srinivas Eeda wrote:
dlm_thread sends a deref message to the master node. At the same time,
another thread sends a new lock request to the master node.
Since dlm_thread
Wengang, thanks for the patch. My comments are inline :)
On 6/8/2010 7:38 AM, Wengang Wang wrote:
dlm_thread(when purges a lockres) races with another thread that is running on
dlmlock_remote(). dlmlock_remote() can add a lock to the blocked list of the
lockres without taking dlm-spinlock.
.
The fix is to add new state DLM_LOCK_RES_IN_USE which would prevent
dlm_thread from purging the lockres and/or unhashing lockres.
Signed-off-by: Srinivas Eeda srinivas.e...@oracle.com
---
fs/ocfs2/dlm/dlmcommon.h |1 +
fs/ocfs2/dlm/dlmlock.c |4
fs/ocfs2/dlm/dlmmaster.c |5
1 - 100 of 146 matches
Mail list logo
