Hi Tiger,
Interesting.
why we never find it in the old reflink test? A little big strange.
Could you please tell me the test case to trigger this issue?
Another thing I want to say is that now your patch removes all the
readahead during CoW. Why? If you want to avoid the
On 08/24/2012 01:33 AM, Sunil Mushran wrote:
On Wed, Aug 22, 2012 at 8:44 PM, Tao Ma t...@tao.ma mailto:t...@tao.ma
wrote:
I guess the final solution will be WRITE_FUA, and I see btrfs uses it to
write out the superblock. It will be handled differently by the
underlying block
Hi Jeff,
On 08/22/2012 10:17 PM, 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
Hi Limin,
Thanks for the report and the patch is already there. Please search the
archive with the subject ocfs2: Fix oops in fallocate().
Thanks
Tao
On 04/13/2012 03:58 PM, Chang Limin wrote:
Hi,
Version linux-3.3
In function
static long ocfs2_fallocate(struct file *file, int
On 01/30/2012 07:44 AM, srinivas eeda wrote:
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?
Hi Srini,
On 01/29/2012 10:13 AM, Srinivas Eeda wrote:
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?
btw, if you are afraid of
over
other requests.
Signed-off-by: Noboru Iwamatsu n_iwama...@jp.fujitsu.com
looks good to me.
Acked-by: Tao Ma boyu...@taobao.com
---
fs/ocfs2/cluster/heartbeat.c |2 +-
1 files changed, 1 insertions(+), 1 deletions(-)
diff --git a/fs/ocfs2/cluster/heartbeat.c b/fs/ocfs2/cluster
On 07/14/2011 03:47 PM, r.giord...@libero.it wrote:
Hello,
this is my scenario:
1)I've created a Pacemaker cluster with the following ocfs package on
opensuse
11.3 64bit
ocfs2console-1.8.0-2.1.x86_64
ocfs2-tools-o2cb-1.8.0-2.1.x86_64
ocfs2-tools-1.8.0-2.1.x86_64
2)I've configured the
Hi wengang,
On 07/12/2011 04:43 PM, Wengang Wang wrote:
When we deleting a direntry from a directory, if it's the first in a block we
invalid it by setting inode to 0; otherwise, we merge the deleted one to the
prior and contiguous direntry. And we don't truncate directories.
There is a
Hi Mark,
On 06/23/2011 05:23 AM, Mark Fasheh wrote:
Fix a corruption that can happen when we have (two or more) outstanding
aio's to an overlapping unaligned region. Ext4
(e9e3bcecf44c04b9e6b505fd8e2eb9cea58fb94d) and xfs recently had to fix
similar issues.
In our case what happens is that
Hi all,
changelog from v1 to v2:
integrated review advices from Sunil.
These are the patches for adding batched discard support in ocfs2. I
have tested it with xfstests 251 and it passed.
btw, I have also run some tests against it(bonnie++, postmark, ffsb and
fs_mark) and there are no
From: Tao Ma boyu...@taobao.com
Add the corresponding ioctl function for FITRIM.
Signed-off-by: Tao Ma boyu...@taobao.com
---
fs/ocfs2/ioctl.c | 24
1 files changed, 24 insertions(+), 0 deletions(-)
diff --git a/fs/ocfs2/ioctl.c b/fs/ocfs2/ioctl.c
index 8f13c59
From: Tao Ma boyu...@taobao.com
Add the corresponding trace event for trim.
Signed-off-by: Tao Ma boyu...@taobao.com
---
fs/ocfs2/alloc.c |7 +++
fs/ocfs2/ocfs2_trace.h | 25 +
2 files changed, 32 insertions(+), 0 deletions(-)
diff --git a/fs/ocfs2
From: Tao Ma boyu...@taobao.com
Add ocfs2_trim_fs to support trimming freed clusters in the
volume. A range will be given and all the freed clusters greater
than minlen will be discarded to the block layer.
Signed-off-by: Tao Ma boyu...@taobao.com
---
fs/ocfs2/alloc.c | 159
Hi Sunil,
Thanks for the review.
On 05/10/2011 07:02 AM, Sunil Mushran wrote:
On 05/06/2011 02:27 AM, Tao Ma wrote:
From: Tao Maboyu...@taobao.com
Add ocfs2_trim_fs to support trimming freed clusters in the
volume. A range will be given and all the freed clusters greater
than minlen
Hi all,
These are the patches for adding batched discard support in ocfs2. I
have tested it with xfstests 251 and it passed.
btw, I have also run some tests against it(bonnie++, postmark, ffsb and
fs_mark) and there are no big difference before and after the discard.
Regards,
Tao
From: Tao Ma boyu...@taobao.com
Add ocfs2_trim_fs to support trimming freed clusters in the
volume. A range will be given and all the freed clusters greater
than minlen will be discarded to the block layer.
Signed-off-by: Tao Ma boyu...@taobao.com
---
fs/ocfs2/alloc.c | 156
From: Tao Ma boyu...@taobao.com
Add the corresponding trace event for trim.
Signed-off-by: Tao Ma boyu...@taobao.com
---
fs/ocfs2/alloc.c |7 +++
fs/ocfs2/ocfs2_trace.h | 25 +
2 files changed, 32 insertions(+), 0 deletions(-)
diff --git a/fs/ocfs2
From: Tao Ma boyu...@taobao.com
Add the corresponding ioctl function for FITRIM.
Signed-off-by: Tao Ma boyu...@taobao.com
---
fs/ocfs2/ioctl.c | 24
1 files changed, 24 insertions(+), 0 deletions(-)
diff --git a/fs/ocfs2/ioctl.c b/fs/ocfs2/ioctl.c
index 8f13c59
Hi Josep,
sorry, I don't subscribed to ocfs2-users after I left Oracle.
On 04/23/2011 08:22 AM, Sunil Mushran wrote:
On 04/22/2011 03:24 PM, Josep Guerrero wrote:
How long did the debugfs output take?
I think about 30 minutes. No more than 50 for sure (just by looking at the
times of
Hi Jeff,
On 03/31/2011 04:28 PM, jeff.liu wrote:
Hello,
Below is a tiny patch for adding cached ACLs support to OCFS2, I have run a
few simple tests like
getacl/setacl/chacl on my single node OCFS2, it works for me.
Your codes seem to not handle the case of cross node update? See node A
On 03/31/2011 04:51 PM, jeff.liu wrote:
Hi Tao,
Tao Ma wrote:
Hi Jeff,
On 03/31/2011 04:28 PM, jeff.liu wrote:
Hello,
Below is a tiny patch for adding cached ACLs support to OCFS2, I have run a
few simple tests like
getacl/setacl/chacl on my single node OCFS2, it works for me.
Your
On 03/10/2011 10:27 AM, Sunil Mushran wrote:
On 03/09/2011 03:01 PM, Mark Fasheh wrote:
On Fri, Mar 04, 2011 at 03:10:33PM +0800, Li Dongyang wrote:
@@ -1870,6 +1870,11 @@ static int __ocfs2_change_file_space(struct file
*file, struct inode *inode,
goto out_inode_unlock;
}
Hi all,
This patch set adds batched discard support to ocfs2. Please check.
Thanks.
Regards,
Tao
___
Ocfs2-devel mailing list
Ocfs2-devel@oss.oracle.com
http://oss.oracle.com/mailman/listinfo/ocfs2-devel
From: Tao Ma boyu...@taobao.com
Add the corresponding ioctl function for FITRIM.
Signed-off-by: Tao Ma boyu...@taobao.com
---
fs/ocfs2/ioctl.c | 24
1 files changed, 24 insertions(+), 0 deletions(-)
diff --git a/fs/ocfs2/ioctl.c b/fs/ocfs2/ioctl.c
index d9bfa90
From: Tao Ma boyu...@taobao.com
Add the corresponding trace event for trim.
Signed-off-by: Tao Ma boyu...@taobao.com
---
fs/ocfs2/alloc.c |7 +++
fs/ocfs2/ocfs2_trace.h | 25 +
2 files changed, 32 insertions(+), 0 deletions(-)
diff --git a/fs/ocfs2
From: Tao Ma boyu...@taobao.com
Add ocfs2_trim_fs to support trimming freed clusters in the
volume. A range will be given and all the freed clusters greater
than minlen will be discarded to the block layer.
Signed-off-by: Tao Ma boyu...@taobao.com
---
fs/ocfs2/alloc.c | 154
On 03/08/2011 12:55 PM, Tristan Ye wrote:
Hi Tao,
Most of codes looks pretty neat to me, few comments inlined below:
Thanks for the review.
Tao Ma wrote:
From: Tao Ma boyu...@taobao.com
Add ocfs2_trim_fs to support trimming freed clusters in the
volume. A range will be given and all
On 03/08/2011 02:23 PM, Tristan Ye wrote:
Tao Ma wrote:
On 03/08/2011 12:55 PM, Tristan Ye wrote:
Hi Tao,
Most of codes looks pretty neat to me, few comments inlined below:
Thanks for the review.
Tao Ma wrote:
From: Tao Ma boyu...@taobao.com
Add ocfs2_trim_fs to support trimming freed
On 03/08/2011 02:53 PM, Tristan Ye wrote:
Tao Ma wrote:
On 03/08/2011 02:23 PM, Tristan Ye wrote:
Tao Ma wrote:
On 03/08/2011 12:55 PM, Tristan Ye wrote:
Hi Tao,
Most of codes looks pretty neat to me, few comments inlined below:
Thanks for the review.
Tao Ma wrote:
From: Tao Ma boyu
On 03/08/2011 12:55 PM, Tristan Ye wrote:
Hi Tao,
Most of codes looks pretty neat to me, few comments inlined below:
Tao Ma wrote:
From: Tao Ma boyu...@taobao.com
Add ocfs2_trim_fs to support trimming freed clusters in the
volume. A range will be given and all the freed clusters
Hi Joel,
On 02/22/2011 05:32 PM, Joel Becker wrote:
Hey everyone,
I just found out from Steven over in GFS2 that there is a kernel
call forget_all_cached_acls(). If a cluster filesystem is using ACL
caching (added by Viro a little while back), they use
forget_all_cached_acls() at lock
Hi Joel,
any plan to integrate this patch set for the next merge window?
Regards,
Tao
On 12/23/2010 03:19 PM, Tao Ma wrote:
Hi all,
This patch set is my attempt to replace our old debug(mlog) with new
trace events.
Wengang Wang has did some work in this field about one year ago.
http
From: Tao Ma boyu...@taobao.com
In cad3f00, ext4_check_dir_entry was modified by adding some unlikely.
Ted described it as This function gets called a lot for large
directories, and the answer is almost always 'no, no, there's no problem'.
This means using unlikely() is a good thing.
ext3 added
On 01/20/2011 09:42 PM, Pawel wrote:
One of ocfs2 filesystem has some errors.
1. fsck.ocfs2 informs me that :I/O error on channel while reading
.. It was NOT TRUE - I was able to read and write entire storage over
the network multiple times.
It means it meet I/O error in some ocfs2
On 01/17/2011 03:08 PM, Tristan Ye wrote:
Tao, any thoughts?
ok, so in general it looks good. some comments are inlined.
Signed-off-by: Tristan Yetristan...@oracle.com
---
fs/ocfs2/move_extents.c | 112
+-
1 files changed, 99 insertions(+),
On 01/05/2011 06:15 AM, Joel Becker wrote:
On Tue, Jan 04, 2011 at 05:06:53PM +0800, Tao Ma wrote:
On 01/01/2011 06:39 AM, Joel Becker wrote:
On Fri, Dec 31, 2010 at 11:11:31PM +0800, Tao Ma wrote:
On 12/31/2010 08:52 PM, Joel Becker wrote:
Overall this seems pretty straightforward
Hi Christoph,
Thanks for the review.
On 12/31/2010 09:15 PM, Christoph Hellwig wrote:
diff --git a/fs/ocfs2/ocfs2_trace.h b/fs/ocfs2/ocfs2_trace.h
index 8d4e49a..2dfc3b0 100644
--- a/fs/ocfs2/ocfs2_trace.h
+++ b/fs/ocfs2/ocfs2_trace.h
@@ -6,6 +6,507 @@
#includelinux/tracepoint.h
On 12/31/2010 07:32 AM, Joel Becker wrote:
On Thu, Dec 16, 2010 at 08:56:06PM +0800, Wengang Wang wrote:
In ocfs2_grab_pages_for_write() code, there is a comment that
1131 if (index == target_index mmap_page) {
1132 /*
1133 *
Hi Tristan,
Some comments inlined.
On 12/28/2010 07:40 PM, Tristan Ye wrote:
snip
+static int ocfs2_move_extents(struct ocfs2_move_extents_context *context)
+{
+ int status;
+ handle_t *handle;
+ struct inode *inode = context-inode;
+ struct buffer_head *di_bh = NULL;
Hi Massimo,
On 12/27/2010 10:31 PM, Massimo Cetra wrote:
Il 03/12/2010 02:22, Tao Ma ha scritto:
I have to say that df is only an approximate way to get the real info
of an ocfs2 volume.
df -i use statfs(2). And since there is no easy way to calculate all
the inodes without
locking every
From: Tao Ma boyu...@taobao.com
This is the 2nd step to remove the debug info of SUPER.
Signed-off-by: Tao Ma boyu...@taobao.com
---
fs/ocfs2/heartbeat.c |4 ++--
fs/ocfs2/ocfs2_trace.h |9 +
2 files changed, 11 insertions(+), 2 deletions(-)
diff --git a/fs/ocfs2/heartbeat.c
From: Tao Ma boyu...@taobao.com
Remove mlog(0) from fs/ocfs2/journal.c and the masklog JOURNAL.
Signed-off-by: Tao Ma boyu...@taobao.com
---
fs/ocfs2/cluster/masklog.c |1 -
fs/ocfs2/cluster/masklog.h |1 -
fs/ocfs2/journal.c | 130 ++--
fs/ocfs2
From: Tao Ma boyu...@taobao.com
Remove mlog(0,...) and mlog(ML_UPTODATE,...) from
fs/ocfs2/uptodate.c and fs/ocfs2/buffer_head_io.c.
The masklog UPTODATE is removed finally.
Signed-off-by: Tao Ma boyu...@taobao.com
---
fs/ocfs2/buffer_head_io.c |3 +-
fs/ocfs2/cluster/masklog.c |1
From: Tao Ma boyu...@taobao.com
Remove mlog(0) from fs/ocfs2/reservations.c and the masklog RESERVATIONS.
Signed-off-by: Tao Ma boyu...@taobao.com
---
fs/ocfs2/cluster/masklog.c |1 -
fs/ocfs2/cluster/masklog.h |1 -
fs/ocfs2/ocfs2_trace.h | 170
From: Tao Ma boyu...@taobao.com
There is no user for masklog AIO, so remove it.
Signed-off-by: Tao Ma boyu...@taobao.com
---
fs/ocfs2/cluster/masklog.c |1 -
fs/ocfs2/cluster/masklog.h |1 -
2 files changed, 0 insertions(+), 2 deletions(-)
diff --git a/fs/ocfs2/cluster/masklog.c b/fs
From: Tao Ma boyu...@taobao.com
This is the 2nd step to remove the debug info of NAMEI.
Signed-off-by: Tao Ma boyu...@taobao.com
---
fs/ocfs2/dir.c | 83 +---
fs/ocfs2/ocfs2_trace.h | 197
2 files changed, 234
From: Tao Ma boyu...@taobao.com
Remove mlog(0) from fs/ocfs2/quota_local.c.
Signed-off-by: Tao Ma boyu...@taobao.com
---
fs/ocfs2/ocfs2_trace.h | 24
fs/ocfs2/quota_local.c | 13 +++--
2 files changed, 31 insertions(+), 6 deletions(-)
diff --git a/fs/ocfs2
From: Tao Ma boyu...@taobao.com
Remove mlog(0) from fs/ocfs2/quota_global.c and
the masklog QUOTA.
Signed-off-by: Tao Ma boyu...@taobao.com
---
fs/ocfs2/cluster/masklog.c |1 -
fs/ocfs2/cluster/masklog.h |1 -
fs/ocfs2/ocfs2_trace.h | 110
From: Tao Ma boyu...@taobao.com
Remove mlog(0) from fs/ocfs2/super.c and the masklog SUPER.
Signed-off-by: Tao Ma boyu...@taobao.com
---
fs/ocfs2/cluster/masklog.c |1 -
fs/ocfs2/cluster/masklog.h |1 -
fs/ocfs2/ocfs2_trace.h | 141
fs
From: Tao Ma boyu...@taobao.com
Remove mlog(0) from fs/ocfs2/xattr.c and the masklog ML_XATTR.
Signed-off-by: Tao Ma boyu...@taobao.com
---
fs/ocfs2/cluster/masklog.c |1 -
fs/ocfs2/cluster/masklog.h |1 -
fs/ocfs2/ocfs2_trace.h | 342
From: Tao Ma boyu...@taobao.com
Remove mlog(0,...) and mlog(ML_BH_IO,...) from
fs/ocfs2/buffer_head_io.c.
The masklog BH_IO is removed finally.
Signed-off-by: Tao Ma boyu...@taobao.com
---
fs/ocfs2/buffer_head_io.c | 37 +++--
fs/ocfs2/cluster/masklog.c |1 -
fs/ocfs2
From: Tao Ma boyu...@taobao.com
Remove mlog(0) from fs/ocfs2/export.c and the masklog EXPORT.
Signed-off-by: Tao Ma boyu...@taobao.com
---
fs/ocfs2/cluster/masklog.c |1 -
fs/ocfs2/cluster/masklog.h |1 -
fs/ocfs2/export.c | 47 +++-
fs/ocfs2/ocfs2_trace.h
Oh, I suddenly found that I send an older version. So it contains some
bugs and I will update them after the review is done.
But it doesn't prevent you guys from review, and please don't run it by
now. ;)
Regards,
Tao
On 12/23/2010 03:19 PM, Tao Ma wrote:
Hi all,
This patch set is my attempt
From: Tao Ma boyu...@taobao.com
When we store system inodes cache in ocfs2_super,
we use a array for global system inodes. But unfortunately,
the range is calculated wrongly which makes it overflow and
pollute ocfs2_super-local_system_inodes.
This patch fix it by setting the range properly
Hi all,
This patch set is my attempt to replace our old debug(mlog) with new
trace events.
Wengang Wang has did some work in this field about one year ago.
http://oss.oracle.com/pipermail/ocfs2-devel/2009-September/005230.html
But at that time we don't have ocfs2 1.6 introduced, so we delay
From: Tao Ma tao...@oracle.com
Tristan Ye has done some refactoring against our truncate
process, so some functions like ocfs2_prepare_truncate and
ocfs2_free_truncate_context are no use and we'd better
remove them.
Signed-off-by: Tao Ma tao...@oracle.com
---
fs/ocfs2/alloc.c | 74
.
Signed-off-by: Wengang Wang wen.gang.w...@oracle.com
Signed-off-by: Tao Ma tao...@oracle.com
---
fs/ocfs2/ocfs2_trace.h | 16
fs/ocfs2/super.c |3 +++
2 files changed, 19 insertions(+), 0 deletions(-)
create mode 100644 fs/ocfs2/ocfs2_trace.h
diff --git a/fs/ocfs2
From: Tao Ma tao...@oracle.com
This is the 2nd step to remove the debug info of DISK_ALLOC.
So this patch removes all mlog(0,...) from localalloc.c and adds
the corresponding tracepoints. Different mlogs have different
solutions.
1. Some are replaced with trace event directly.
2. Some
From: Tao Ma tao...@oracle.com
This is the first try of replacing debug mlog(0,...) to
trace events. Wengang has did some work in his original
patch
http://oss.oracle.com/pipermail/ocfs2-devel/2009-November/005513.html
But he didn't finished it.
So this patch removes all mlog(0,...) from alloc.c
From: Tao Ma tao...@oracle.com
Since all 4 files, localalloc.c, suballoc.c, alloc.c and
resize.c, which use DISK_ALLOC are changed to trace events,
Remove masklog DISK_ALLOC totally.
Signed-off-by: Tao Ma tao...@oracle.com
---
fs/ocfs2/alloc.c |1 -
fs/ocfs2/cluster/masklog.c
From: Tao Ma tao...@oracle.com
This is the 3rd step to remove the debug info of DISK_ALLOC.
Signed-off-by: Tao Ma tao...@oracle.com
---
fs/ocfs2/ocfs2_trace.h | 329 +++-
fs/ocfs2/suballoc.c| 107
2 files changed, 381 insertions
From: Tao Ma tao...@oracle.com
mlog_exit is used to record the exit status of a function.
But because it is added in so many functions, if we enable it,
the system logs get filled up quickly and cause too much I/O.
So actually no one can open it for a production system or even
for a test
From: Tao Ma tao...@oracle.com
Change all the mlog(0, in fs/ocfs2/refcounttree.c to trace events.
And finally remove masklog ML_REFCOUNT.
Signed-off-by: Tao Ma tao...@oracle.com
---
fs/ocfs2/cluster/masklog.c |1 -
fs/ocfs2/cluster/masklog.h |1 -
fs/ocfs2/ocfs2_trace.h | 430
forget to say.
you can view from
http://repo.or.cz/w/taoma-kernel.git/shortlog/refs/heads/mlog_replace
and pull from
git://repo.or.cz/taoma-kernel.git mlog_replace
Regards,
Tao
On 12/23/2010 03:19 PM, Tao Ma wrote:
Hi all,
This patch set is my attempt to replace our old debug(mlog) with new
From: Tao Ma boyu...@taobao.com
ocfs2_iget is used to get/create inode. Only iget5_locked
will give us an inode = NULL. So move this check ahead of
ocfs2_read_locked_inode so that we don't need to check
inode before we read and unlock inode. This is also helpful
for trace event(see the next patch
From: Tao Ma tao...@oracle.com
Change all the mlog(0, in fs/ocfs2/mmap.c to trace events.
And finally remove masklog FILE_IO.
Signed-off-by: Tao Ma tao...@oracle.com
---
fs/ocfs2/cluster/masklog.c |1 -
fs/ocfs2/cluster/masklog.h |1 -
fs/ocfs2/mmap.c|7 +++
fs/ocfs2
From: Tao Ma boyu...@taobao.com
Remove mlog(0) from fs/ocfs2/extent_map.c and the masklog EXTENT_MAP.
Signed-off-by: Tao Ma boyu...@taobao.com
---
fs/ocfs2/cluster/masklog.c |1 -
fs/ocfs2/cluster/masklog.h |1 -
fs/ocfs2/extent_map.c |5 ++---
fs/ocfs2/ocfs2_trace.h | 32
From: Tao Ma boyu...@taobao.com
As there are no such debug information in fs/ocfs2/ioctl.c,
fs/ocfs2/locks.c and fs/ocfs2/sysfile.c, ML_INODE are also
removed.
Signed-off-by: Tao Ma boyu...@taobao.com
---
fs/ocfs2/acl.c |1 -
fs/ocfs2/ioctl.c |1 -
fs/ocfs2/locks.c |1 -
3 files
From: Tao Ma tao...@oracle.com
Remove all the mlog(0, in fs/ocfs2/aops.c.
Signed-off-by: Tao Ma tao...@oracle.com
---
fs/ocfs2/aops.c| 55
fs/ocfs2/ocfs2_trace.h | 168
2 files changed, 197 insertions(+), 26 deletions
From: Tao Ma boyu...@taobao.com
Remove mlog(0) from fs/ocfs2/inode.c and the masklog INODE.
Signed-off-by: Tao Ma boyu...@taobao.com
---
fs/ocfs2/cluster/masklog.c |1 -
fs/ocfs2/cluster/masklog.h |1 -
fs/ocfs2/inode.c | 107 +
fs/ocfs2/ocfs2_trace.h
From: Tao Ma boyu...@taobao.com
This is the 2nd step to remove the debug info of INODE.
Signed-off-by: Tao Ma boyu...@taobao.com
---
fs/ocfs2/file.c| 190 -
fs/ocfs2/ocfs2_trace.h | 367
fs/ocfs2/sysfile.c
From: Tao Ma boyu...@taobao.com
In ocfs2_double_lock, when ocfs2_inode_lock for inode1 fails, we
just unlock inode2 and return without releasing buffer we get from
inode_lock(inode2). The good thing is that it is freed by the only
caller ocfs2_rename when it exits.
But I don't think
From: Tao Ma boyu...@taobao.com
When we set/clear the dyn_features for an inode we hold the ip_lock.
So do it when we set/clear OCFS2_INDEXED_DIR_FL also.
Signed-off-by: Tao Ma boyu...@taobao.com
---
fs/ocfs2/dir.c |4
1 files changed, 4 insertions(+), 0 deletions(-)
diff --git a/fs
On 12/10/2010 01:33 PM, Sunil Mushran wrote:
ok. But this patch is a bug fix for 2.6.37. So is still needed.
I don't mean this bug fix isn't needed. Sorry if I described it wrongly.
What I want to say is that with my new patch set, these type of things
would happen again. ;)
Regards,
Tao
On
Hi Mark,
On 12/07/2010 09:13 AM, Mark Fasheh wrote:
On Fri, Nov 26, 2010 at 02:35:58PM +0800, Tao Ma wrote:
Hi Joel,
On 11/25/2010 06:08 PM, Joel Becker wrote:
On Wed, Oct 20, 2010 at 02:08:17PM +0800, Tao Ma wrote:
j_trans_barrier in ocfs2 is used to protect some journal operations
Hi Massimo,
Massimo Cetra wrote:
Il 09/11/2010 02:06, Tao Ma ha scritto:
Hi Massimo,
From the output, it seems that you shouldn't meet with ENOSPC problem
at that time. So when you meet with ENOSPC, could you please do 'sync'
first(that will sync the metadata to the disk so that stat_sysdir
Hi Massimo,
I just noticed anther thing.
The inode_alloc: isn't full, so node1 should steal inode from node0,
why it can't? You delete some files in node0?
Regards,
Tao
On 12/03/2010 09:22 AM, Tao Ma wrote:
Hi Massimo,
Massimo Cetra wrote:
Il 09/11/2010 02:06, Tao Ma ha scritto:
Hi
Hi Tristan,
Your patch forgets to set the new flag properly in direct read.
Regards,
Tao
On 11/29/2010 03:54 PM, Tristan Ye wrote:
Due to newly-introduced 'coherency=full' O_DIRECT writes also takes the EX
rw_lock like buffered writes did(rw_level == 1), it turns out messing the
usage
Hi Tristan,
On 11/29/2010 05:21 PM, Tristan Ye wrote:
Due to newly-introduced 'coherency=full' O_DIRECT writes also takes the EX
rw_lock like buffered writes did(rw_level == 1), it turns out messing the
usage of 'level' in ocfs2_dio_end_io() up, which caused i_alloc_sem being
failed to get
Hi Joel,
On 11/25/2010 06:08 PM, Joel Becker wrote:
On Wed, Oct 20, 2010 at 02:08:17PM +0800, Tao Ma wrote:
j_trans_barrier in ocfs2 is used to protect some journal operations
in ocfs2. So normally, it is used as belows:
1. In journal transaction. When we start a transaction, We
Hi Tristan,
On 11/22/2010 10:22 AM, Tristan Ye wrote:
Tao Ma wrote:
Hi Tristan,
Just add joel to the cc in case he has a different option.
On 11/19/2010 04:38 PM, Tristan Ye wrote:
Former logic of ocfs2_file_aio_write() was a bit stricky to unlock
the rw_lock
and i_alloc_sem, by using
Hi Tristan,
Just add joel to the cc in case he has a different option.
On 11/19/2010 04:38 PM, Tristan Ye wrote:
Former logic of ocfs2_file_aio_write() was a bit stricky to unlock the rw_lock
and i_alloc_sem, by using some private bits in struct 'iocb' to communite with
: Goldwyn Rodrigues rgold...@suse.de
Signed-off-by: Tao Ma tao...@oracle.com
---
fs/ocfs2/ocfs2.h |6 +++---
1 files changed, 3 insertions(+), 3 deletions(-)
diff --git a/fs/ocfs2/ocfs2.h b/fs/ocfs2/ocfs2.h
index d840821..1efea36 100644
--- a/fs/ocfs2/ocfs2.h
+++ b/fs/ocfs2/ocfs2.h
@@ -159,7
Joel Becker wrote:
On Wed, Nov 10, 2010 at 10:52:20AM +0800, Tao Ma wrote:
ENTRY is used to record the entry of a function.
But because it is added in so many functions, if we enable it,
the system logs get filled up quickly and cause too much I/O.
So actually no one can open
Joel Becker wrote:
On Wed, Nov 10, 2010 at 10:52:21AM +0800, Tao Ma wrote:
mlog_exit is used to record the exit status of a function.
But because it is added in so many functions, if we enable it,
the system logs get filled up quickly and cause too much I/O.
So actually no one can open
Joel Becker wrote:
On Wed, Nov 10, 2010 at 10:52:22AM +0800, Tao Ma wrote:
About one year ago, Wengang Wang tried some first steps
to add tracepoints to ocfs2. But at that time we don't
have ocfs2 1.6 introduced, so we delay this work until now
after ocfs2 1.6 and UEK(Unbreakable
Joel Becker wrote:
On Wed, Nov 10, 2010 at 10:52:23AM +0800, Tao Ma wrote:
So this patch removes all mlog(0,...) from alloc.c and adds
the corresponding tracepoints. Different mlogs have different
solutions.
1. Some are replaced with trace event directly.
2. Some are replaced and some
Hi all,
This patch set is my first try to add trace event to ocfs2.
Wengang Wang has did some work in this field about one year ago.
http://oss.oracle.com/pipermail/ocfs2-devel/2009-September/005230.html
But at that time we don't have ocfs2 1.6 introduced, so we delay this
work until now after
...@oracle.com
Signed-off-by: Tao Ma tao...@oracle.com
---
fs/ocfs2/ocfs2_trace.h | 16
fs/ocfs2/super.c |3 +++
2 files changed, 19 insertions(+), 0 deletions(-)
create mode 100644 fs/ocfs2/ocfs2_trace.h
diff --git a/fs/ocfs2/ocfs2_trace.h b/fs/ocfs2/ocfs2_trace.h
new
return value, it is replaced with
mlog(0,...).
mlog_exit_ptr is changed to mlog(0.
Signed-off-by: Tao Ma tao...@oracle.com
---
fs/ocfs2/alloc.c | 26 +++
fs/ocfs2/aops.c | 19 ++-
fs/ocfs2/buffer_head_io.c|7 ++--
fs/ocfs2/cluster
.
What's more, it defines some event classes so that we can use
them later.
Signed-off-by: Wengang Wang wen.gang.w...@oracle.com
Signed-off-by: Tao Ma tao...@oracle.com
---
fs/ocfs2/alloc.c | 165 +
fs/ocfs2/ocfs2_trace.h | 485
Tristan Ye has done some refactoring against our truncate
process, so some functions like ocfs2_prepare_truncate and
ocfs2_free_truncate_context are no use and we'd better
remove them.
Signed-off-by: Tao Ma tao...@oracle.com
---
fs/ocfs2/alloc.c | 74
Hi Massimo,
On 11/08/2010 11:14 PM, Massimo Cetra wrote:
Hi all,
i've been using OCFS2 for a while (2 years or so, now) but i still have
a big problem.
on each filesystem that i have, it happens that, even if df reports that
the disk has plenty of free space avalaible, no one can write any
at that point and try again. This method
is inspired by Mark Fasheh mfas...@suse.com. Thanks.
Cc: Mark Fasheh mfas...@suse.com
Signed-off-by: Tao Ma tao...@oracle.com
---
fs/ocfs2/alloc.c |3 ++
fs/ocfs2/aops.c | 59 +-
fs/ocfs2/ocfs2.h |5
On 11/04/2010 09:46 AM, Joel Becker wrote:
On Tue, Oct 26, 2010 at 03:29:08PM +0800, Tao Ma wrote:
Recently, one of our colleagues meet with a problem that if we
write/delete a 32mb files repeatly, we will get an ENOSPC in
the end. And the corresponding bug is 1288.
http://oss.oracle.com
at that point and try again. This method
is inspired by Mark Fasheh mfas...@suse.com. Thanks.
Cc: Mark Fasheh mfas...@suse.com
Signed-off-by: Tao Ma tao...@oracle.com
---
fs/ocfs2/alloc.c |3 ++
fs/ocfs2/aops.c | 59 +-
fs/ocfs2/ocfs2.h |2 +
3
On 10/26/2010 04:28 PM, tristan wrote:
Hi Tao,
Just some tiny comments;)
Tao Ma wrote:
Recently, one of our colleagues meet with a problem that if we
write/delete a 32mb files repeatly, we will get an ENOSPC in
the end. And the corresponding bug is 1288.
http://oss.oracle.com/bugzilla
Hi Joel,
Joel Becker wrote:
On Mon, Oct 11, 2010 at 04:46:39PM +0800, Tristan Ye wrote:
Currently, default behavior of O_DIRECT writes was allowing
concurrent writing among nodes, no cluster coherency guaranteed
(no EX locks was taken), it hurts buffered reads on other nodes
by reading
Tao Ma wrote:
Hi Joel,
Joel Becker wrote:
On Mon, Oct 11, 2010 at 04:46:39PM +0800, Tristan Ye wrote:
Currently, default behavior of O_DIRECT writes was allowing
concurrent writing among nodes, no cluster coherency guaranteed
(no EX locks was taken), it hurts buffered reads
1 - 100 of 688 matches
Mail list logo