regression in 2.6.37?

regression in 2.6.37?

[Jiro SEKIBA]

Hi,

I found that none of the checpoints are flagged 'i'.
It looks like it happens after 2.6.37.

So whenever I only execute "chcp ss <cp>" without any filesystem change,
nilfs volume create new cp without 'i'.
Before 2.6.37, it creates new cp with 'i'.

It this somewhat new behavior or regression?

thanks,

regards
-- 
Jiro SEKIBA <jir-hfpbi5WX9J54Eiagz67IpQ@public.gmane.org>
--
To unsubscribe from this list: send the line "unsubscribe linux-nilfs" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: regression in 2.6.37?

[Jiro SEKIBA]

At Mon, 28 Feb 2011 10:33:24 +0900,
Jiro SEKIBA wrote:
> 
> Hi,
> 
> I found that none of the checpoints are flagged 'i'.
> It looks like it happens after 2.6.37.
> 
> So whenever I only execute "chcp ss <cp>" without any filesystem change,
> nilfs volume create new cp without 'i'.
> Before 2.6.37, it creates new cp with 'i'.

This, by the way, causes that nilfs2_ss_manager creates
snapshots forever and useless as a result.

So I would really like to turn it back to the normal behavior.
It is really critical for the application.

thanks

regards,

> It this somewhat new behavior or regression?
> 
> thanks,
> 
> regards
> -- 
> Jiro SEKIBA <jir-hfpbi5WX9J54Eiagz67IpQ@public.gmane.org>
> --
> To unsubscribe from this list: send the line "unsubscribe linux-nilfs" in
> the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html
> 
> 
> 



-- 
Jiro SEKIBA <jir-hfpbi5WX9J54Eiagz67IpQ@public.gmane.org>
--
To unsubscribe from this list: send the line "unsubscribe linux-nilfs" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: regression in 2.6.37?

[Ryusuke Konishi]

On Mon, 28 Feb 2011 10:58:36 +0900, Jiro SEKIBA wrote:
> 
> At Mon, 28 Feb 2011 10:33:24 +0900,
> Jiro SEKIBA wrote:
> > 
> > Hi,
> > 
> > I found that none of the checpoints are flagged 'i'.
> > It looks like it happens after 2.6.37.
> > 
> > So whenever I only execute "chcp ss <cp>" without any filesystem change,
> > nilfs volume create new cp without 'i'.
> > Before 2.6.37, it creates new cp with 'i'.
> 
> This, by the way, causes that nilfs2_ss_manager creates
> snapshots forever and useless as a result.
> 
> So I would really like to turn it back to the normal behavior.
> It is really critical for the application.
> 
> thanks
> 
> regards,

Arrh, that sounds so bad.  Ok, I'll look into the problem ASAP.

Ryusuke Konishi

> > It this somewhat new behavior or regression?
> > 
> > thanks,
> > 
> > regards
> > -- 
> > Jiro SEKIBA <jir-hfpbi5WX9J54Eiagz67IpQ@public.gmane.org>
--
To unsubscribe from this list: send the line "unsubscribe linux-nilfs" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: regression in 2.6.37?

[Ryusuke Konishi]

On Mon, 28 Feb 2011 11:21:46 +0900 (JST), Ryusuke Konishi wrote:
> On Mon, 28 Feb 2011 10:58:36 +0900, Jiro SEKIBA wrote:
> > 
> > At Mon, 28 Feb 2011 10:33:24 +0900,
> > Jiro SEKIBA wrote:
> > > 
> > > Hi,
> > > 
> > > I found that none of the checpoints are flagged 'i'.
> > > It looks like it happens after 2.6.37.
> > > 
> > > So whenever I only execute "chcp ss <cp>" without any filesystem change,
> > > nilfs volume create new cp without 'i'.
> > > Before 2.6.37, it creates new cp with 'i'.
> > 
> > This, by the way, causes that nilfs2_ss_manager creates
> > snapshots forever and useless as a result.
> > 
> > So I would really like to turn it back to the normal behavior.
> > It is really critical for the application.
> > 
> > thanks
> > 
> > regards,
> 
> Arrh, that sounds so bad.  Ok, I'll look into the problem ASAP.
> 
> Ryusuke Konishi

Actually, this tourned out to be a regression on 2.6.37.
The following patch will fix this.

Could you confirm that it corrects your problem?

Thanks,
Ryusuke Konishi
---
From: Ryusuke Konishi <konishi.ryusuke-Zyj7fXuS5i5L9jVzuh4AOg@public.gmane.org>

nilfs2: fix regression that i-flag is not set on changeless checkpoints

According to the report from Jiro SEKIBA titled "regression in
2.6.37?"  (Message-Id: <8739n8vs1f.wl%jir-27yqGEOhnJbQT0dZR+AlfA@public.gmane.org>), on 2.6.37 and
later kernels, lscp command no longer displays "i" flag on checkpoints
that snapshot operations or garbage collection created.

This is a regression of nilfs2 checkpointing function, and it's
critical since it broke behavior of a part of nilfs2 applications.
For instance, snapshot manager of TimeBrowse gets to create
meaningless snapshots continuously; snapshot creation triggers another
checkpoint, but applications cannot distinguish whether the new
checkpoint contains meaningful changes or not without the i-flag.

This patch fixes the regression and brings that application behavior
back to normal.

Reported-by: Jiro SEKIBA <jir-hfpbi5WX9J54Eiagz67IpQ@public.gmane.org>
Signed-off-by: Ryusuke Konishi <konishi.ryusuke-Zyj7fXuS5i5L9jVzuh4AOg@public.gmane.org>
Cc: stable <stable-DgEjT+Ai2ygdnm+yROfE0A@public.gmane.org>  [2.6.37]
---
 fs/nilfs2/segment.c |    3 ++-
 1 files changed, 2 insertions(+), 1 deletions(-)

diff --git a/fs/nilfs2/segment.c b/fs/nilfs2/segment.c
index 55ebae5..2de9f63 100644
--- a/fs/nilfs2/segment.c
+++ b/fs/nilfs2/segment.c
@@ -430,7 +430,8 @@ static void nilfs_segctor_begin_finfo(struct nilfs_sc_info *sci,
 	nilfs_segctor_map_segsum_entry(
 		sci, &sci->sc_binfo_ptr, sizeof(struct nilfs_finfo));
 
-	if (inode->i_sb && !test_bit(NILFS_SC_HAVE_DELTA, &sci->sc_flags))
+	if (NILFS_I(inode)->i_root &&
+	    !test_bit(NILFS_SC_HAVE_DELTA, &sci->sc_flags))
 		set_bit(NILFS_SC_HAVE_DELTA, &sci->sc_flags);
 	/* skip finfo */
 }
-- 
1.7.3.5

--
To unsubscribe from this list: send the line "unsubscribe linux-nilfs" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: regression in 2.6.37?

[Jiro SEKIBA]

At Mon, 28 Feb 2011 13:52:56 +0900 (JST),
Ryusuke Konishi wrote:
> 
> On Mon, 28 Feb 2011 11:21:46 +0900 (JST), Ryusuke Konishi wrote:
> > On Mon, 28 Feb 2011 10:58:36 +0900, Jiro SEKIBA wrote:
> > > 
> > > At Mon, 28 Feb 2011 10:33:24 +0900,
> > > Jiro SEKIBA wrote:
> > > > 
> > > > Hi,
> > > > 
> > > > I found that none of the checpoints are flagged 'i'.
> > > > It looks like it happens after 2.6.37.
> > > > 
> > > > So whenever I only execute "chcp ss <cp>" without any filesystem change,
> > > > nilfs volume create new cp without 'i'.
> > > > Before 2.6.37, it creates new cp with 'i'.
> > > 
> > > This, by the way, causes that nilfs2_ss_manager creates
> > > snapshots forever and useless as a result.
> > > 
> > > So I would really like to turn it back to the normal behavior.
> > > It is really critical for the application.
> > > 
> > > thanks
> > > 
> > > regards,
> > 
> > Arrh, that sounds so bad.  Ok, I'll look into the problem ASAP.
> > 
> > Ryusuke Konishi
> 
> Actually, this tourned out to be a regression on 2.6.37.
> The following patch will fix this.
> 
> Could you confirm that it corrects your problem?

Thank you for the prompt reaction!
I confirmed the patch worked fine with 2.6.38-rc6.

thanks again

regards,

> Thanks,
> Ryusuke Konishi
> ---
> From: Ryusuke Konishi <konishi.ryusuke-Zyj7fXuS5i5L9jVzuh4AOg@public.gmane.org>
> 
> nilfs2: fix regression that i-flag is not set on changeless checkpoints
> 
> According to the report from Jiro SEKIBA titled "regression in
> 2.6.37?"  (Message-Id: <8739n8vs1f.wl%jir-27yqGEOhnJbQT0dZR+AlfA@public.gmane.org>), on 2.6.37 and
> later kernels, lscp command no longer displays "i" flag on checkpoints
> that snapshot operations or garbage collection created.
> 
> This is a regression of nilfs2 checkpointing function, and it's
> critical since it broke behavior of a part of nilfs2 applications.
> For instance, snapshot manager of TimeBrowse gets to create
> meaningless snapshots continuously; snapshot creation triggers another
> checkpoint, but applications cannot distinguish whether the new
> checkpoint contains meaningful changes or not without the i-flag.
> 
> This patch fixes the regression and brings that application behavior
> back to normal.
> 
> Reported-by: Jiro SEKIBA <jir-hfpbi5WX9J54Eiagz67IpQ@public.gmane.org>
> Signed-off-by: Ryusuke Konishi <konishi.ryusuke-Zyj7fXuS5i5L9jVzuh4AOg@public.gmane.org>
> Cc: stable <stable-DgEjT+Ai2ygdnm+yROfE0A@public.gmane.org>  [2.6.37]
> ---
>  fs/nilfs2/segment.c |    3 ++-
>  1 files changed, 2 insertions(+), 1 deletions(-)
> 
> diff --git a/fs/nilfs2/segment.c b/fs/nilfs2/segment.c
> index 55ebae5..2de9f63 100644
> --- a/fs/nilfs2/segment.c
> +++ b/fs/nilfs2/segment.c
> @@ -430,7 +430,8 @@ static void nilfs_segctor_begin_finfo(struct nilfs_sc_info *sci,
>  	nilfs_segctor_map_segsum_entry(
>  		sci, &sci->sc_binfo_ptr, sizeof(struct nilfs_finfo));
>  
> -	if (inode->i_sb && !test_bit(NILFS_SC_HAVE_DELTA, &sci->sc_flags))
> +	if (NILFS_I(inode)->i_root &&
> +	    !test_bit(NILFS_SC_HAVE_DELTA, &sci->sc_flags))
>  		set_bit(NILFS_SC_HAVE_DELTA, &sci->sc_flags);
>  	/* skip finfo */
>  }
> -- 
> 1.7.3.5
> 
> --
> To unsubscribe from this list: send the line "unsubscribe linux-nilfs" in
> the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html
> 
> 
> 


-- 
Jiro SEKIBA <jir-hfpbi5WX9J54Eiagz67IpQ@public.gmane.org>
--
To unsubscribe from this list: send the line "unsubscribe linux-nilfs" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html