Users on the list recently complained about differences across
filesystems w.r.t. how to mount without a journal replay.
In the discussion it was noted that xfs's "norecovery" option is
perhaps more descriptively accurate than "noload," so let's make
that an alias for ext3.
Also show this status in /proc/mounts
Signed-off-by: Eric Sandeen <[email protected]>
---
diff --git a/Documentation/filesystems/ext3.txt b/Documentation/filesystems/ext3.txt
index 05d5cf1..8b8d973 100644
--- a/Documentation/filesystems/ext3.txt
+++ b/Documentation/filesystems/ext3.txt
@@ -33,7 +33,7 @@ journal_dev=devnum When the external journal device's major/minor numbers
in devnum.
noload Don't load the journal on mounting. Note that this forces
- mount of inconsistent filesystem, which can lead to
+norecovery mount of inconsistent filesystem, which can lead to
various problems.
data=journal All data are committed into the journal prior to being
diff --git a/fs/ext3/super.c b/fs/ext3/super.c
index 427496c..c14fb95 100644
--- a/fs/ext3/super.c
+++ b/fs/ext3/super.c
@@ -636,6 +636,9 @@ static int ext3_show_options(struct seq_file *seq, struct vfsmount *vfs)
if (test_opt(sb, DATA_ERR_ABORT))
seq_puts(seq, ",data_err=abort");
+ if (test_opt(sb, NOLOAD))
+ seq_puts(seq, ",norecovery");
+
ext3_show_quota_options(seq, sb);
return 0;
@@ -818,6 +821,7 @@ static const match_table_t tokens = {
{Opt_reservation, "reservation"},
{Opt_noreservation, "noreservation"},
{Opt_noload, "noload"},
+ {Opt_noload, "norecovery"},
{Opt_nobh, "nobh"},
{Opt_bh, "bh"},
{Opt_commit, "commit=%u"},
On Mon 16-11-09 16:50:49, Eric Sandeen wrote:
> Users on the list recently complained about differences across
> filesystems w.r.t. how to mount without a journal replay.
>
> In the discussion it was noted that xfs's "norecovery" option is
> perhaps more descriptively accurate than "noload," so let's make
> that an alias for ext3.
>
> Also show this status in /proc/mounts
>
> Signed-off-by: Eric Sandeen <[email protected]>
OK, provided ext4 will have the same change, I'm fine with the patch.
So how is it with ext4?
Honza
> ---
>
> diff --git a/Documentation/filesystems/ext3.txt b/Documentation/filesystems/ext3.txt
> index 05d5cf1..8b8d973 100644
> --- a/Documentation/filesystems/ext3.txt
> +++ b/Documentation/filesystems/ext3.txt
> @@ -33,7 +33,7 @@ journal_dev=devnum When the external journal device's major/minor numbers
> in devnum.
>
> noload Don't load the journal on mounting. Note that this forces
> - mount of inconsistent filesystem, which can lead to
> +norecovery mount of inconsistent filesystem, which can lead to
> various problems.
>
> data=journal All data are committed into the journal prior to being
> diff --git a/fs/ext3/super.c b/fs/ext3/super.c
> index 427496c..c14fb95 100644
> --- a/fs/ext3/super.c
> +++ b/fs/ext3/super.c
> @@ -636,6 +636,9 @@ static int ext3_show_options(struct seq_file *seq, struct vfsmount *vfs)
> if (test_opt(sb, DATA_ERR_ABORT))
> seq_puts(seq, ",data_err=abort");
>
> + if (test_opt(sb, NOLOAD))
> + seq_puts(seq, ",norecovery");
> +
> ext3_show_quota_options(seq, sb);
>
> return 0;
> @@ -818,6 +821,7 @@ static const match_table_t tokens = {
> {Opt_reservation, "reservation"},
> {Opt_noreservation, "noreservation"},
> {Opt_noload, "noload"},
> + {Opt_noload, "norecovery"},
> {Opt_nobh, "nobh"},
> {Opt_bh, "bh"},
> {Opt_commit, "commit=%u"},
>
--
Jan Kara <[email protected]>
SUSE Labs, CR
Jan Kara wrote:
> On Mon 16-11-09 16:50:49, Eric Sandeen wrote:
>> Users on the list recently complained about differences across
>> filesystems w.r.t. how to mount without a journal replay.
>>
>> In the discussion it was noted that xfs's "norecovery" option is
>> perhaps more descriptively accurate than "noload," so let's make
>> that an alias for ext3.
>>
>> Also show this status in /proc/mounts
>>
>> Signed-off-by: Eric Sandeen <[email protected]>
> OK, provided ext4 will have the same change, I'm fine with the patch.
> So how is it with ext4?
I did send the same patch for ext4
"[PATCH 2/2] ext4: make "norecovery" an alias for "noload""
Thanks,
-Eric
> Honza
>> ---
>>
>> diff --git a/Documentation/filesystems/ext3.txt b/Documentation/filesystems/ext3.txt
>> index 05d5cf1..8b8d973 100644
>> --- a/Documentation/filesystems/ext3.txt
>> +++ b/Documentation/filesystems/ext3.txt
>> @@ -33,7 +33,7 @@ journal_dev=devnum When the external journal device's major/minor numbers
>> in devnum.
>>
>> noload Don't load the journal on mounting. Note that this forces
>> - mount of inconsistent filesystem, which can lead to
>> +norecovery mount of inconsistent filesystem, which can lead to
>> various problems.
>>
>> data=journal All data are committed into the journal prior to being
>> diff --git a/fs/ext3/super.c b/fs/ext3/super.c
>> index 427496c..c14fb95 100644
>> --- a/fs/ext3/super.c
>> +++ b/fs/ext3/super.c
>> @@ -636,6 +636,9 @@ static int ext3_show_options(struct seq_file *seq, struct vfsmount *vfs)
>> if (test_opt(sb, DATA_ERR_ABORT))
>> seq_puts(seq, ",data_err=abort");
>>
>> + if (test_opt(sb, NOLOAD))
>> + seq_puts(seq, ",norecovery");
>> +
>> ext3_show_quota_options(seq, sb);
>>
>> return 0;
>> @@ -818,6 +821,7 @@ static const match_table_t tokens = {
>> {Opt_reservation, "reservation"},
>> {Opt_noreservation, "noreservation"},
>> {Opt_noload, "noload"},
>> + {Opt_noload, "norecovery"},
>> {Opt_nobh, "nobh"},
>> {Opt_bh, "bh"},
>> {Opt_commit, "commit=%u"},
>>
On Thu 19-11-09 09:57:44, Eric Sandeen wrote:
> Jan Kara wrote:
> > On Mon 16-11-09 16:50:49, Eric Sandeen wrote:
> >> Users on the list recently complained about differences across
> >> filesystems w.r.t. how to mount without a journal replay.
> >>
> >> In the discussion it was noted that xfs's "norecovery" option is
> >> perhaps more descriptively accurate than "noload," so let's make
> >> that an alias for ext3.
> >>
> >> Also show this status in /proc/mounts
> >>
> >> Signed-off-by: Eric Sandeen <[email protected]>
> > OK, provided ext4 will have the same change, I'm fine with the patch.
> > So how is it with ext4?
>
> I did send the same patch for ext4
>
> "[PATCH 2/2] ext4: make "norecovery" an alias for "noload""
Obviously I'm too lame at googling ;). OK, I've merged the patch to
for_next branch.
Honza
--
Jan Kara <[email protected]>
SUSE Labs, CR
On Thu, Nov 19, 2009 at 04:12:33PM +0100, Jan Kara wrote:
> On Mon 16-11-09 16:50:49, Eric Sandeen wrote:
> > Users on the list recently complained about differences across
> > filesystems w.r.t. how to mount without a journal replay.
> >
> > In the discussion it was noted that xfs's "norecovery" option is
> > perhaps more descriptively accurate than "noload," so let's make
> > that an alias for ext3.
> >
> > Also show this status in /proc/mounts
> >
> > Signed-off-by: Eric Sandeen <[email protected]>
> OK, provided ext4 will have the same change, I'm fine with the patch.
> So how is it with ext4?
Queued for the next merge window. I did slightly change the
documentation in Documentation/filesystems/ext4.txt to make
"norecovery" be the canonnical option name, since that's what Eric had
show up in /proc/mounts. That seems fair, we might as well try to add
consistency between different Linux filesystems.
- Ted
[email protected] wrote:
> On Thu, Nov 19, 2009 at 04:12:33PM +0100, Jan Kara wrote:
>> On Mon 16-11-09 16:50:49, Eric Sandeen wrote:
>>> Users on the list recently complained about differences across
>>> filesystems w.r.t. how to mount without a journal replay.
>>>
>>> In the discussion it was noted that xfs's "norecovery" option is
>>> perhaps more descriptively accurate than "noload," so let's make
>>> that an alias for ext3.
>>>
>>> Also show this status in /proc/mounts
>>>
>>> Signed-off-by: Eric Sandeen <[email protected]>
>> OK, provided ext4 will have the same change, I'm fine with the patch.
>> So how is it with ext4?
>
> Queued for the next merge window. I did slightly change the
> documentation in Documentation/filesystems/ext4.txt to make
> "norecovery" be the canonnical option name, since that's what Eric had
> show up in /proc/mounts. That seems fair, we might as well try to add
> consistency between different Linux filesystems.
>
> - Ted
FWIW, the nilfs folks just added an option to do the same thing, and
I steered them towards "norecovery" - we may just get some consistency,
yet!
-Eric