This patch fixes the use of XATTR_NODE_OFFSET.
o The offset should not use several MSB bits which are used by marking node
blocks.
o IS_DNODE should handle XATTR_NODE_OFFSET to avoid potential abnormality
during the fsync call.
Signed-off-by: Jaegeuk Kim
---
fs/f2fs/f2fs.h | 12
fsync when a
file has a modified xattr node block.
This approach is able to degrade the performance, but normally the checkpoint
overhead is shown at the initial fsync call after the xattr entry changes.
Once the checkpoint is done, no additional overhead would be occurred.
Signed-off-by: J
gt; + sbi->n_orphans--;
NAK.
We should not release them here.
Only f2fs_evict_inode can do it.
> }
> if (!page)
> goto end;
--
Jaegeuk Kim
Samsung
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message t
Hi Al,
2013-07-19 (금), 08:49 +0100, Al Viro:
> On Fri, Jul 19, 2013 at 12:40:47PM +0900, Kim Jaegeuk wrote:
> > Hi,
> >
> > 2013. 7. 18. 6:22?? "Al Viro" :
> > >
> > > On Thu, Jul 18, 2013 at 06:11:23PM +0900, Jaegeu
As similar as the i_pino fix, i_name also should be fixed when i_nlink is 1.
The errorneous scenario is like this.
1. touch test1
2. link test1 test2
3. unlink test2
4. fsync test1
After this, i_name should be test1.
CC: Al Viro
Signed-off-by: Jaegeuk Kim
---
fs/f2fs/file.c | 10
Let's check the free space in prior to the main process of allocating a new node
page.
Signed-off-by: Jaegeuk Kim
---
fs/f2fs/node.c | 18 +-
1 file changed, 9 insertions(+), 9 deletions(-)
diff --git a/fs/f2fs/node.c b/fs/f2fs/node.c
index bb8fbda..858a333 100644
--- a/fs
'
At this moment, the checkpoint stores inconsistent data where the inode has
i_xattr_nid but actual xattr node block is not allocated yet.
So, we should assign the real i_xattr_nid only after its xattr node block is
allocated.
Signed-off-by: Jaegeuk Kim
---
fs/f2fs/node.c | 3 +++
f
_GC all the time.
Thanks,
> continue;
> if (IS_CURSEC(sbi, GET_SECNO(sbi, segno)))
> continue;
--
Jaegeuk Kim
Samsung
signature.asc
Description: This is a digitally signed message part
page.
> f2fs_delete_entry(de, page, inode);
>
> /* In order to evict this inode, we set it dirty */
--
Jaegeuk Kim
Samsung
signature.asc
Description: This is a digitally signed message part
We should not change the on-disk layout.
Instead, simply we can deal with it by changing original condition check
like below.
---
From ccc2546eded1efd2d6ed98f8aee7d7ce247cb4a2 Mon Sep 17 00:00:00 2001
From: Jaegeuk Kim
Date: Sun, 3 Mar 2013 13:58:05 +0900
Subject: [PATCH] f2fs: align f2fs
[] do_notify_resume+0x71/0xb0
[] int_signal+0x12/0x17
Reported-by: Chris Fries
Signed-off-by: Jaegeuk Kim
---
fs/f2fs/node.c | 2 ++
1 file changed, 2 insertions(+)
diff --git a/fs/f2fs/node.c b/fs/f2fs/node.c
index 3df43b4..9641534 100644
--- a/fs/f2fs/node.c
+++ b/fs/f2fs/node.c
@@ -1492,6
order to avoid that, this patch fixes the lock order of node and data
blocks in which the node block lock is covered by the data block lock.
Signed-off-by: Jaegeuk Kim
---
fs/f2fs/data.c | 17 +++--
1 file changed, 11 insertions(+), 6 deletions(-)
diff --git a/fs/f2fs/data.c b/fs/f2fs
; > inc_valid_inode_count(sbi);
> >
> > --
> > 1.8.1.3.566.gaa39828
> >
> > --
> > To unsubscribe from this list: send the line "unsubscribe linux-fsdevel" in
> > the body of a message to majord...@vger.kernel.org
> > More majordomo info at http://vger.kernel.org/majordomo-info.html
> >
--
Jaegeuk Kim
Samsung
signature.asc
Description: This is a digitally signed message part
2013-05-15 (수), 13:04 +0900, Namjae Jeon:
> 2013/5/15, Jaegeuk Kim :
> > In get_lock_data_page, if there is a data race between get_dnode_of_data
> > for
> > node and grab_cache_page for data, f2fs is able to face with the following
> > BUG_ON(dn.data_blkaddr == NEW_A
if (!((long int)ofs % (NIDS_PER_BLOCK + 1)))
Great catch!
I'll apply this.
Thank you.
> return false;
> }
> return true;
--
Jaegeuk Kim
Samsung
signature.asc
Description: This is a digitally signed message part
In order to be aware of prefree and free sections during FG_GC, let's start with
write_checkpoint().
Signed-off-by: Jaegeuk Kim
---
fs/f2fs/gc.c | 4 +++-
1 file changed, 3 insertions(+), 1 deletion(-)
diff --git a/fs/f2fs/gc.c b/fs/f2fs/gc.c
index 83cec8f..37b05e1 100644
--- a/fs/f2fs
struct f2fs_sb_info *sbi = F2FS_SB(inode->i_sb);
> +
> + f2fs_balance_fs(sbi);
>
> blk_start = pg_start << PAGE_CACHE_SHIFT;
> blk_end = pg_end << PAGE_CACHE_SHIFT;
> -- 1.8.0
--
Jaegeuk Kim
Samsung
signature.asc
Description: This is a digitally signed message part
Hrycay
> Hi,
> With the latest commit 9995bf953a83749abd9fa22f72ab2b0be341025a
> About introducing the global locking method in ‘f2fs’,
> I think we no longer will have a case of deadlock happening in this path.
Hi, Namjae.
I found that this bug still exists in the new locking model.
Please see the v3 patch. :)
Thanks,
>
> Thanks.
--
Jaegeuk Kim
Samsung
signature.asc
Description: This is a digitally signed message part
Chang log from v2:
o resolve a deadlock path
[ilock count][call path]
1
0 f2fs_unlink
0- truncate_hole
0 - try to grab ilock (wait)
From 66b15368995b7218a37f589b9ada0c63794c4dc6 Mon Sep 17 00:00:00 2001
From: Jaegeuk Kim
ize)
f2fs_write_begin [enter] (dev,ino,pos,len,flags)
7. patch 7/7:
write_checkpoint_before_blockop [before block_operations()]
(dev,is_mount,"try to block operations")
write_checkpoint_after_blockop [after block_operations()]
(dev,is_mount,"done block operations")
write_checkpoint_exit [exit] (dev,is_mount,"done checkpoint")
>
> >
> > Acked-by: Steven Rostedt
> >
> > -- Steve
> >
> >
> >
--
Jaegeuk Kim
Samsung
signature.asc
Description: This is a digitally signed message part
change log from v2:
o slighly change the previous approach
From 42e49df33bacafc20434e39fc127c9722c9f8691 Mon Sep 17 00:00:00 2001
From: Jaegeuk Kim
Date: Tue, 30 Apr 2013 11:33:27 +0900
Subject: [PATCH] f2fs: avoid deadlock during evict after f2fs_gc
Cc: linux-fsde...@vger.kernel.org, linux
> 2 * MAX_FREE_NIDS)
> + break;
Could you explain when this can happen?
IMO, this is an unnecessary condition check, since the below condition
that includes FREE_NID_PAGES already limits the number of free nids.
Thanks,
> if (i++ == FREE_NID_
04474c06f319 Mon Sep 17 00:00:00 2001
From: Haicheng Li
Date: Mon, 6 May 2013 23:15:43 +0800
Subject: [PATCH] f2fs: optimize scan_nat_page()
Cc: linux-fsde...@vger.kernel.org, linux-kernel@vger.kernel.org,
linux-f2fs-de...@lists.sourceforge.net
When nm_i->fcnt > 2 * MAX_FREE_NIDS, stop scanning o
, &nm_i->free_nid_list, list).
But, this is not covered by free_nid_list_lock, resulting in null pointer bug.
This patch moves this checking routine inside add_free_nid() in order not to use
the spin_lock.
Signed-off-by: Jaegeuk Kim
---
fs/f2fs/node.c | 34 ++
ff-by: Jaegeuk Kim
---
fs/f2fs/f2fs.h | 2 --
fs/f2fs/namei.c | 15
fs/f2fs/node.c | 76 +
fs/f2fs/node.h | 6 -
fs/f2fs/xattr.c | 2 --
5 files changed, 17 insertions(+), 84 deletions(-)
diff --git a/fs/f2fs/f2fs.h b/fs
Sorry, please ignore this patch.
2013-05-08 (수), 09:56 +0900, Jaegeuk Kim:
> This mechanism revealed two issues: bug and performance.
>
> Now, iput() doesn't guarantee that the inode is freed completely due to the
> linked f2fs_drop_inode().
> So, in the case of failure on
build_free_nids()
f2fs: optimize scan_nat_page()
Jaegeuk Kim (35):
f2fs: fix to unlock node page when it was truncated
f2fs: read with READ_SYNC when getting dnode page
f2fs: introduce readahead mode of node pages
f2fs: align f2fs maximum name length to linux based filesy
2013-05-08 (수), 14:24 +0800, Haicheng Li:
> On Tue, May 07, 2013 at 07:33:59PM +0900, Jaegeuk Kim wrote:
> > 2013-05-06 (월), 23:15 +0800, Haicheng Li:
> > > diff --git a/fs/f2fs/node.c b/fs/f2fs/node.c
> > > index 1fe3fe2..3136224 100644
> > > --- a/fs/f2
ix inconsistent using of NM_WOUT_THRESHOLD
f2fs: remove useless #include as we're now using
sysfs as debug entry.
f2fs: bugfix for alloc_nid_failed()
f2fs: code cleanup for scan_nat_page() and build_free_nids()
f2fs: optimize scan_nat_page()
Jaegeuk Kim (35):
tch as above your comment ?
I can handle it. :)
Thanks,
>
> Thanks.
> >>
> >>> + rw |= REQ_META;
> >>> +
> >>> if (sbi->bio[btype]) {
> >>> struct bio_private *p = sbi->bio[btype]->bi_private;
> >>>
,16 R () 1521688 + 144 [0]
: block_rq_complete: 8,16 R () 1522128 + 192 [0]
: block_rq_complete: 8,16 R () 1523256 + 328 [0]
...
Note that this issue should be addressed in checkpoint, and some readahead
flows too.
Signed-off-by: Jaegeuk Kim
---
fs/f2fs/checkpoint.c | 5 +
fs/f2fs/d
If there is no victim segments selected by background GC, let's wait
a little bit longer time to collect dirty segments.
By default, let's give 10 minutes.
Signed-off-by: Jaegeuk Kim
---
fs/f2fs/gc.h | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/fs/f2fs/gc.h
Hi, Namjae,
Agreed. How about this?
Chang log from v1:
o change timings - min 30s, max 60s, nogc 5 min
o remove nonreachable routine
o consider NOGC_SLEEP_TIME in increate/decrease_sleep_time
From 806e344624414fcf9fc87f6193265859027d51b5 Mon Sep 17 00:00:00 2001
From: Jaegeuk Kim
Date: Wed
It is more obvious that add_free_nid checks whether the free nid is zero or not.
Signed-off-by: Jaegeuk Kim
---
fs/f2fs/node.c | 8
1 file changed, 4 insertions(+), 4 deletions(-)
diff --git a/fs/f2fs/node.c b/fs/f2fs/node.c
index a0aa044..c8f48d4 100644
--- a/fs/f2fs/node.c
+++ b/fs
urnal nat entries, and nat cache entries.
We should consider carefullly not to serve free nids intermediately made by
build_free_nids.
We can get stable free nids only after build_free_nids is done.
Signed-off-by: Jaegeuk Kim
---
fs/f2fs/f2fs.h | 2 +-
fs/f2fs/node.c
This can help when debugging the free nid allocation flows.
Signed-off-by: Jaegeuk Kim
---
fs/f2fs/namei.c | 3 ++-
include/trace/events/f2fs.h | 7 +++
2 files changed, 9 insertions(+), 1 deletion(-)
diff --git a/fs/f2fs/namei.c b/fs/f2fs/namei.c
index c57fd18..4aa26e5 100644
) << PAGE_CACHE_SHIFT) > i_size_read(inode)) {
> @@ -90,7 +88,9 @@ static int f2fs_vm_page_mkwrite(struct vm_area_struct *vma,
> set_page_dirty(page);
> SetPageUptodate(page);
>
> - file_update_time(vma->vm_file);
> +mapped:
> + /* fill the page
We call lock_page when we need to update a page after readpage.
Between grab and lock page, the page can be truncated by other thread.
So, we should check the page after lock_page whether it was truncated or not.
Signed-off-by: Jaegeuk Kim
---
fs/f2fs/checkpoint.c | 4
fs/f2fs/data.c
k
The above three threads are able to incur deadlock condition.
In order to address this issue, let's use f2fs_iget_nowait again in f2fs_gc so
that thread 2 doesn't need to wait for inode->i_lock.
Signed-off-by: Jaegeuk Kim
---
fs/f2fs/f2fs.h | 1 +
fs/f2fs/gc.c| 2 +-
f
bio_get_nr_vecs, so this patch replace it
with max_hw_blocks using queue_max_sectors.
Signed-off-by: Jaegeuk Kim
---
fs/f2fs/node.c| 6 ++
fs/f2fs/segment.c | 2 +-
fs/f2fs/segment.h | 11 +++
3 files changed, 14 insertions(+), 5 deletions(-)
diff --git a/fs/f2fs/node.c b/fs/f2fs
change log from v1:
o found another deadlock case, so change the approach
From 145b6ee0b2523bad0e28c09f4ac5bb050cd22085 Mon Sep 17 00:00:00 2001
From: Jaegeuk Kim
Date: Tue, 30 Apr 2013 11:33:27 +0900
Subject: [PATCH] f2fs: avoid deadlock during evict after f2fs_gc
Cc: linux-fsde
The allocated page used by the recovery is not on HIGHMEM, so that we don't
need to use kmap/kunmap.
Signed-off-by: Jaegeuk Kim
---
fs/f2fs/recovery.c | 8 +++-
1 file changed, 3 insertions(+), 5 deletions(-)
diff --git a/fs/f2fs/recovery.c b/fs/f2fs/recovery.c
index 774efdb..5f
If we got an error after lock_page, we should unlock it before exit.
Signed-off-by: Jaegeuk Kim
---
fs/f2fs/recovery.c | 5 ++---
1 file changed, 2 insertions(+), 3 deletions(-)
diff --git a/fs/f2fs/recovery.c b/fs/f2fs/recovery.c
index 5fcdd08..edb89a3 100644
--- a/fs/f2fs/recovery.c
+++ b/fs
This patch is for passing a locked node page to get_dnode_of_data.
Signed-off-by: Jaegeuk Kim
---
fs/f2fs/data.c | 12 +++-
fs/f2fs/dir.c | 4 ++--
fs/f2fs/f2fs.h | 2 +-
fs/f2fs/file.c | 2 +-
4 files changed, 11 insertions(+), 9 deletions(-)
diff --git a/fs/f2fs/data.c b/fs/f2fs
This patch adds some trivial debugging messages in the recovery process.
Signed-off-by: Jaegeuk Kim
---
fs/f2fs/node.c | 1 -
fs/f2fs/recovery.c | 44 +---
2 files changed, 25 insertions(+), 20 deletions(-)
diff --git a/fs/f2fs/node.c b/fs/f2fs
whole recovery process.
And then, write_checkpoint flushes all the dirty dentry blocks, and nicely we
can put the stale dir inodes from the dirty_dir_inode_list.
Signed-off-by: Jaegeuk Kim
---
fs/f2fs/checkpoint.c | 23 +++
fs/f2fs/f2fs.h | 2 ++
fs/f2fs/recovery.c
curs the zero-valued i_pino during the
recovery.
So, this patch modifies the creation flow to fix the synchronization order of
inode page with its inode.
Signed-off-by: Jaegeuk Kim
---
fs/f2fs/data.c | 1 +
fs/f2fs/dir.c | 85 +++---
fs/f2fs/f2f
If get_dnode_of_data gets a locked node page, let's skip redundant
get_node_page calls.
This is for the futher enhancement.
Signed-off-by: Jaegeuk Kim
---
fs/f2fs/node.c | 9 ++---
1 file changed, 6 insertions(+), 3 deletions(-)
diff --git a/fs/f2fs/node.c b/fs/f2fs/node.c
index 96
The reason of using sbi->por_doing is to alleviate data writes during the
recovery.
The find_fsync_dnodes() produces some dirty dentry pages, so we should
cover it too with sbi->por_doing.
Signed-off-by: Jaegeuk Kim
---
fs/f2fs/recovery.c | 4 ++--
1 file changed, 2 insertions(+), 2 del
This por_doing check is totally not related to the recovery process.
Signed-off-by: Jaegeuk Kim
---
fs/f2fs/namei.c | 3 +--
1 file changed, 1 insertion(+), 2 deletions(-)
diff --git a/fs/f2fs/namei.c b/fs/f2fs/namei.c
index 47abc97..729b285 100644
--- a/fs/f2fs/namei.c
+++ b/fs/f2fs/namei.c
We don't need to assign a value redundantly.
Signed-off-by: Jaegeuk Kim
---
fs/f2fs/recovery.c | 5 ++---
1 file changed, 2 insertions(+), 3 deletions(-)
diff --git a/fs/f2fs/recovery.c b/fs/f2fs/recovery.c
index 60c8a50..2941987 100644
--- a/fs/f2fs/recovery.c
+++ b/fs/f2fs/recov
If an inode is recovered with its dentry, it will not invoke __f2fs_add_link,
since the recovery routine checks its dentry before calling __f2fs_add_link.
Signed-off-by: Jaegeuk Kim
---
fs/f2fs/recovery.c | 6 +-
1 file changed, 1 insertion(+), 5 deletions(-)
diff --git a/fs/f2fs
Sorry for the noise.
This patch should be ignored.
Thanks,
2013-05-20 (월), 12:32 +0900, Jaegeuk Kim:
> If an inode is recovered with its dentry, it will not invoke __f2fs_add_link,
> since the recovery routine checks its dentry before calling __f2fs_add_link.
>
> Signed-off-by:
s_truncate+0x5c/0xa0
[] SyS_truncate+0xe/0x10
[] system_call_fastpath+0x16/0x1b
Signed-off-by: Jaegeuk Kim
---
fs/f2fs/data.c | 4 ++--
fs/f2fs/inode.c | 3 +--
2 files changed, 3 insertions(+), 4 deletions(-)
diff --git a/fs/f2fs/data.c b/fs/f2fs/data.c
index c320f7f..1644fff 100644
---
sue.
Signed-off-by: Jaegeuk Kim
---
fs/f2fs/inode.c | 6 --
1 file changed, 6 deletions(-)
diff --git a/fs/f2fs/inode.c b/fs/f2fs/inode.c
index a18946e..b44a4c1 100644
--- a/fs/f2fs/inode.c
+++ b/fs/f2fs/inode.c
@@ -109,12 +109,6 @@ struct inode *f2fs_iget(struct super_block *sb, unsigned
long
Introduce a simple macro function for readability.
Signed-off-by: Jaegeuk Kim
---
fs/f2fs/f2fs.h | 5 +
fs/f2fs/file.c | 2 +-
fs/f2fs/super.c | 2 +-
3 files changed, 7 insertions(+), 2 deletions(-)
diff --git a/fs/f2fs/f2fs.h b/fs/f2fs/f2fs.h
index 9182b27..6594ce1 100644
--- a/fs/f2fs
ion check for the reuse of locked direct node
pages prior to the get_node_page call.
Signed-off-by: Jaegeuk Kim
---
fs/f2fs/f2fs.h | 1 +
fs/f2fs/file.c | 2 +-
fs/f2fs/recovery.c | 26 +-
3 files changed, 23 insertions(+), 6 deletions(-)
diff --git a/fs/f2fs/f2f
This patch adds error handling codes of check_index_in_prev_nodes and its
caller, do_recover_data.
Signed-off-by: Jaegeuk Kim
---
fs/f2fs/recovery.c | 24 +++-
1 file changed, 15 insertions(+), 9 deletions(-)
diff --git a/fs/f2fs/recovery.c b/fs/f2fs/recovery.c
index
If -ENOSPC is met during f2fs_link, we should not make the inode as bad.
The inode is still alive.
Signed-off-by: Jaegeuk Kim
---
fs/f2fs/namei.c | 1 -
1 file changed, 1 deletion(-)
diff --git a/fs/f2fs/namei.c b/fs/f2fs/namei.c
index 729b285..71aa305 100644
--- a/fs/f2fs/namei.c
+++ b/fs
Use the following helper function committed by Al.
commit 7de9c6ee3ecffd99e1628e81a5ea5468f7581a1f
Author: Al Viro
Date: Sat Oct 23 11:11:40 2010 -0400
new helper: ihold()
...
Signed-off-by: Jaegeuk Kim
---
fs/f2fs/namei.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff
lock[i]);
> mutex_init(&sbi->node_write);
> sbi->por_doing = 0;
> + sbi->forced_fg = 0;
> spin_lock_init(&sbi->stat_lock);
> init_rwsem(&sbi->bio_sem);
> init_sb_info(sbi);
> @@ -691,6 +826,14 @@ static int f2fs_fill_super(struct super_block *sb, void
> *data, int silent)
> "the device does not support discard");
> }
>
> + sbi->s_kobj.kset = f2fs_kset;
> + init_completion(&sbi->s_kobj_unregister);
> + err = kobject_init_and_add(&sbi->s_kobj, &f2fs_ktype, NULL,
> + "%s", sb->s_id);
> + if (err)
> + goto fail;
> +
> +
> return 0;
> fail:
> stop_gc_thread(sbi);
> @@ -765,6 +908,9 @@ static int __init init_f2fs_fs(void)
> err = create_checkpoint_caches();
> if (err)
> goto fail;
> + f2fs_kset = kset_create_and_add("f2fs", NULL, fs_kobj);
> + if (!f2fs_kset)
> + goto fail;
> err = register_filesystem(&f2fs_fs_type);
> if (err)
> goto fail;
> @@ -781,6 +927,7 @@ static void __exit exit_f2fs_fs(void)
> destroy_gc_caches();
> destroy_node_manager_caches();
> destroy_inodecache();
> + kset_unregister(f2fs_kset);
> }
>
> module_init(init_f2fs_fs)
--
Jaegeuk Kim
Samsung
signature.asc
Description: This is a digitally signed message part
; The most notable one is PostgreSQL pgbench v8.4.11
> Transactions per second: 3032 in kernel 3.9
> down to 832 in kernel 3.10-rc1
>
> On Wed, May 8, 2013 at 2:10 PM, Jaegeuk Kim wrote:
> > Hi Linus,
> >
> > I've rebased one of patches, so could you consider the foll
If there remains some unwritten blocks from the recovery, we should not call
iput on that directory inode.
Otherwise, we can loose some dentry blocks after the recovery.
Signed-off-by: Jaegeuk Kim
---
fs/f2fs/checkpoint.c | 7 ---
1 file changed, 4 insertions(+), 3 deletions(-)
diff --git
and its inode
and then recover newly created dentry.
Signed-off-by: Jaegeuk Kim
---
fs/f2fs/recovery.c | 23 ---
1 file changed, 16 insertions(+), 7 deletions(-)
diff --git a/fs/f2fs/recovery.c b/fs/f2fs/recovery.c
index 0dd2ce1..85f5e54 100644
--- a/fs/f2fs/recovery.c
+++ b
- write_chekcpoint
If *dir*'s i_count is not 1 (i.e., root dir), remove_dirty_dir_inode is called
later and then iput is triggered again due to the FI_DELAY_IPUT flag.
So, let's unset the flag properly once iput is triggered.
Signed-off-by: Jaegeuk Kim
---
fs/f2fs/ch
Hi,
2013-05-27 (월), 13:45 +0900, Namjae Jeon:
> 2013/5/27, Jaegeuk Kim :
> > Hi Namjae,
> Hi Jaegeuk.
>
> First, Thanks for your interest.
> >
> > This is an interesting functionality.
> > Could you describe why and when we need to do this?
> > What
2013-03-26 (화), 09:48 +0900, Namjae Jeon:
> 2013/3/25, Jaegeuk Kim :
> > This patch removes data_version check flow during the fsync call.
> > The original purpose for the use of data_version was to avoid writng inode
> > pages redundantly by the fsync calls repeatedly.
>
2013-03-27 (수), 09:57 +0900, Namjae Jeon:
> 2013/3/27, Jaegeuk Kim :
> > 2013-03-26 (화), 09:48 +0900, Namjae Jeon:
> >> 2013/3/25, Jaegeuk Kim :
> >> > This patch removes data_version check flow during the fsync call.
> >> > The original purpose for t
Let's use a macro to get the total number of sections.
Signed-off-by: Jaegeuk Kim
---
fs/f2fs/debug.c | 7 +++
fs/f2fs/segment.c | 19 +--
fs/f2fs/segment.h | 1 +
3 files changed, 13 insertions(+), 14 deletions(-)
diff --git a/fs/f2fs/debug.c b/fs/f2fs/debug.c
The foreground GCs are triggered under not enough free sections.
So, we should not skip moving valid blocks in the victim segments.
Signed-off-by: Jaegeuk Kim
---
fs/f2fs/gc.c | 46 ++
1 file changed, 34 insertions(+), 12 deletions(-)
diff --git a/fs
.
However, the SSR mode selects victims freely from any log types, which can set
multiple bits across the various bitmap types.
So, this patch eliminates this inconsistency.
Signed-off-by: Jaegeuk Kim
---
fs/f2fs/segment.c | 23 ++-
1 file changed, 18 insertions(+), 5 deletions
In order to do GC more reliably, I'd like to lock the vicitm summary page
until its GC is completed, and also prevent any checkpoint process.
Signed-off-by: Jaegeuk Kim
---
fs/f2fs/gc.c| 8 +---
fs/f2fs/node.c | 2 +-
fs/f2fs/super.c | 7 +--
3 files changed, 7 insertions(+
This patch adds a new condition that allocates free segments in the current
active section even if SSR is needed.
Otherwise, f2fs cannot allocate remained free segments in the section since
SSR finds dirty segments only.
Signed-off-by: Jaegeuk Kim
---
fs/f2fs/segment.c | 13 +
1
A macro should not use duplicate parameter names.
Signed-off-by: Jaegeuk Kim
---
fs/f2fs/segment.h | 14 +++---
1 file changed, 7 insertions(+), 7 deletions(-)
diff --git a/fs/f2fs/segment.h b/fs/f2fs/segment.h
index e399bd4..c0d7740 100644
--- a/fs/f2fs/segment.h
+++ b/fs/f2fs
When allocating a new segment under the LFS mode, we should keep the section
boundary.
Signed-off-by: Jaegeuk Kim
---
fs/f2fs/segment.c | 3 ++-
1 file changed, 2 insertions(+), 1 deletion(-)
diff --git a/fs/f2fs/segment.c b/fs/f2fs/segment.c
index 179a13e..b3486f3 100644
--- a/fs/f2fs
ground GCs, since
there is no need to read victim blocks during foreground GCs.
By the fact that the foreground GC reclaims segments in a section unit, it'd
be better to manage this bitmap based on the section granularity.
Signed-off-by: Jaegeuk Kim
---
fs/f2fs/checkpoint.c |
In get_node_page, we do not need to call lock_page all the time.
If the node page is cached as uptodate,
1. grab_cache_page locks the page,
2. read_node_page unlocks the page, and
3. lock_page is called for further process.
Let's avoid this.
Signed-off-by: Jaegeuk Kim
---
fs/f2fs/node.c
2013-03-04 (월), 15:10 +0900, Namjae Jeon:
> 2013/3/3, Jaegeuk Kim :
> > 2013-03-02 (토), 12:41 +0900, Namjae Jeon:
> >> From: Namjae Jeon
> >>
> >> Actual dirty of pages will occur in f2fs_delete_entry so move the
> >> f2fs_balance_fs just before d
2013-03-04 (월), 15:25 +0900, Namjae Jeon:
> 2013/3/3, Jaegeuk Kim :
> > We should not change the on-disk layout.
> > Instead, simply we can deal with it by changing original condition check
> > like below.
>
> Even though the changes you suggested are ok. But there is o
Hi,
2013-01-18 (금), 16:29 +0300, Dan Carpenter:
> Hello Jaegeuk Kim,
>
> The patch 7bc0900347e0: "f2fs: add garbage collection functions" from
> Nov 2, 2012, has an off-by-one bug.
>
>429 block_t start_bidx_of_node(unsigned int node_ofs)
>430 {
n. And also, as a cleanup, I've converted it to use
> > list_for_each_entry instead of list_for_each.
> >
> > Signed-off-by: Dan Carpenter
> > ---
> > Static analysis stuff. Untested. Please review carefully.
>
> Makes sense to me.
>
> Revi
e cleanups
Dan Carpenter (1):
f2fs: use _safe() version of list_for_each
Huajun Li (2):
f2fs: update f2fs document to reflect SIT/NAT layout correctly
f2fs: update f2fs partition info about SIT/NAT layout
Jaegeuk Kim (10):
f2fs: initialize newly allocated dnode
2012-10-12 (금), 16:30 +0400, Vyacheslav Dubeyko:
> On Wed, 2012-10-10 at 18:43 +0900, Jaegeuk Kim wrote:
> > [snip]
> > > > How about the following scenario?
> > > > 1. data "a" is newly written.
> > > > 2. checkpoint "A" is done.
&g
]);
> ckpt->alloc_type[i + CURSEG_HOT_NODE] =
> curseg_alloc_type(sbi, i + CURSEG_HOT_NODE);
> }
>
> which will take and drop that same lock 3 times in quick succession, and then
> do it again for 3 other locks (And th
hat when that last block is written and so calls
>complete(p->wait);
> then all the blocks have been written.
> I don't think that is a valid assumption in general. Various things can
> re-order blocks. Back when we had BIO_BARRIER, a barrier request would force
&g
a summaries.
Note that this scenario is not from the umount procedure.
Basically f2fs writes three data summary blocks for current active logs
inside the checkpoint pack.
And NAT and SIT journal entries are stored in hot and cold data summary
blocks.
So, if the CP_COMPACT_SUM_FLAG is not set, f2fs
nm_i, ne);
> > + }
> > + write_unlock(&nm_i->nat_tree_lock);
> > + return nr_shrink;
> > +}
>
>
> This code looks wrong to me, in a small way.
> The return value is only ever tested for whether it is zero or not.
> For that last 'retu
message to majord...@vger.kernel.org
> More majordomo info at http://vger.kernel.org/majordomo-info.html
> Please read the FAQ at http://www.tux.org/lkml/
--
Jaegeuk Kim
Samsung
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majord...@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
e enhanced by
cooperating with a readahead mechanism in VFS.
Thanks,
>
> Thanks.
--
Jaegeuk Kim
Samsung
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majord...@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
S
> > +manages a bitmap. Each bit represents the validity of a block, and the
> > bitmap is
> > +composed of a bit stream covering whole blocks in main area.
>
> Do you take the type of data into account in the cleaner? In particular, do
> you
> try to group blocks
time from the same defaults, but with the option of overriding it
> using
> a debugfs tool.
>
Good point!
I'll think about a user-made list.
Thanks,
> Arnd
> --
> To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
> the bo
nd the line "unsubscribe linux-kernel" in
> > the body of a message to majord...@vger.kernel.org
> > More majordomo info at http://vger.kernel.org/majordomo-info.html
> > Please read the FAQ at http://www.tux.org/lkml/
>
> --
> To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
> the body of a message to majord...@vger.kernel.org
> More majordomo info at http://vger.kernel.org/majordomo-info.html
> Please read the FAQ at http://www.tux.org/lkml/
--
Jaegeuk Kim
Samsung
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majord...@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
> On Sun, Oct 14, 2012 at 03:19:37PM +, Arnd Bergmann wrote:
> > On Sunday 14 October 2012, Vyacheslav Dubeyko wrote:
> > > On Oct 14, 2012, at 11:09 AM, Jaegeuk Kim wrote:
> > > > 2012-10-14 (일), 02:21 +0400, Vyacheslav Dubeyko:
> > > Extended attribut
> On Monday 15 October 2012, Changman Lee wrote:
> > 2012년 10월 15일 월요일에 Arnd Bergmann님이 작성:
> > > It is only a performance hint though, so it is not a correctness issue the
> > > file system gets it wrong. In order to do efficient garbage collection, a
> > > log
> > > structured file system should
2012-10-16 (화), 11:38 +, Arnd Bergmann:
> On Tuesday 16 October 2012, Jaegeuk Kim wrote:
> > On Monday 15 October 2012, Dave Chinner wrote:
> > > On Sun, Oct 14, 2012 at 03:19:37PM +, Arnd Bergmann wrote:
> > > > On Sunday 14 October 2012, Vyacheslav Dube
2012-10-16 (화), 16:14 +, Arnd Bergmann:
> On Tuesday 16 October 2012, Jaegeuk Kim wrote:
> > Thank you for a lot of points to be addressed. :)
> > Maybe it's time to summarize them.
> > Please let me know what I misunderstood.
> >
> > [In v2]
>
's why
> xattrs have different namespaces like "system", "security" and
> "user". Only user attributes are truly for user data
> - the rest are for communicating information to the kernel
>
I agree that "system" is used
> On Wed, Oct 17, 2012 at 07:30:21AM +0900, Jaegeuk Kim wrote:
> > > > > OTOH, I think xattr itself is for users, not for communicating
> > > > > between file system and users.
> > > >
> > > > No, you are mistaken in that point, as Dave exp
> 2012-10-16 (화), 16:14 +, Arnd Bergmann:
> > On Tuesday 16 October 2012, Jaegeuk Kim wrote:
> > > Thank you for a lot of points to be addressed. :)
> > > Maybe it's time to summarize them.
> > > Please let me know what I misunderstood.
> > >
&
ranch
f2fs
3. Contact
Jaegeuk Kim
4. Short log
Greg Kroah-Hartman (1):
f2fs: move proc files to debugfs
Jaegeuk Kim (16):
f2fs: add document
f2fs: add on-disk layout
f2fs: add superblock and major in-memory structure
f2fs: add super block operations
f2fs: add checkpoint operations
201 - 300 of 2588 matches
Mail list logo
