Re: [PATCH 12/19] libext2fs: Bump default number of reserved inodes to 64

[Jan Kara <jack@suse.cz>]

On Fri 07-08-15 13:11:05, Andreas Dilger wrote:
> On Aug 7, 2015, at 4:51 AM, Jan Kara <jack@suse.com> wrote:
> > 
> > From: Jan Kara <jack@suse.cz>
> > 
> > We ran out of reserved inodes so bump the default number of reserved
> > inodes to 64 to get some breathing space. Otherwise we have to do a full
> > fs scan when increasing number of reserved inodes when some feature
> > needing another reserved inode is enabled. This consumes 13.5 KB on a
> > filesystem which is negligible these days.
> 
> This patch also needs to fix all of the regression tests that will
> fail because the inode numbers are different.  Otherwise "make check"
> will fail.

Yeah, it was a bit tedious work but now all the tests are passing. Thanks
for noticing.

								Honza

> 
> Cheers, Andreas
> 
> > Signed-off-by: Jan Kara <jack@suse.cz>
> > ---
> > lib/ext2fs/initialize.c | 2 +-
> > misc/mke2fs.8.in        | 8 ++++----
> > misc/mke2fs.conf.5.in   | 7 ++++---
> > 3 files changed, 9 insertions(+), 8 deletions(-)
> > 
> > diff --git a/lib/ext2fs/initialize.c b/lib/ext2fs/initialize.c
> > index 75fbf8ee0061..0ecf4606ce48 100644
> > --- a/lib/ext2fs/initialize.c
> > +++ b/lib/ext2fs/initialize.c
> > @@ -186,7 +186,7 @@ errcode_t ext2fs_initialize(const char *name, int flags,
> > 
> > 	set_field(s_rev_level, EXT2_GOOD_OLD_REV);
> > 	if (super->s_rev_level >= EXT2_DYNAMIC_REV) {
> > -		set_field(s_first_ino, EXT2_GOOD_OLD_FIRST_INO);
> > +		set_field(s_first_ino, 64);
> > 		set_field(s_inode_size, EXT2_GOOD_OLD_INODE_SIZE);
> > 		if (super->s_inode_size >= sizeof(struct ext2_inode_large)) {
> > 			int extra_isize = sizeof(struct ext2_inode_large) -
> > diff --git a/misc/mke2fs.8.in b/misc/mke2fs.8.in
> > index 520a07185f9f..6eef1d97bfbf 100644
> > --- a/misc/mke2fs.8.in
> > +++ b/misc/mke2fs.8.in
> > @@ -386,10 +386,10 @@ Do not attempt to discard blocks at mkfs time.
> > @QUOTA_MAN_COMMENT@behavior is to initialize both user and group quotas.
> > .TP
> > .BI reserved_inodes= number
> > -Specify the number of inodes reserved for system files. This number must be
> > -at least 10. Currently 10 is enough but future features may require additional
> > -reserved inodes. Reserving more inodes after file system is created requires
> > -full file system scan so it can take a long time.
> > +Specify the number of inodes reserved for system files. This number must be at
> > +least 10, default is 64. Currently 10 is enough but future features may require
> > +additional reserved inodes.
> 
> I'm not sure it is worthwhile to mention "currently 10 is enough",
> since this should be updated every time some new inode is reserved,
> but it will likely be forgotten.  I don't think users really care
> about this in the end.
> 
> > Reserving more inodes after file system is created
> > +requires full file system scan so it can take a long time.
> > .RE
> > .TP
> > .BI \-f " fragment-size"
> > diff --git a/misc/mke2fs.conf.5.in b/misc/mke2fs.conf.5.in
> > index 06ca9e4eabc4..b35767bfedf2 100644
> > --- a/misc/mke2fs.conf.5.in
> > +++ b/misc/mke2fs.conf.5.in
> > @@ -197,9 +197,10 @@ reserved ratio. This value can be a floating point number.
> > .TP
> > .I reserved_inodes
> > This relation specifies the default number of inodes reserved for system files.
> > -The number must be at least 10. Currently 10 is enough but future features may
> > -require additional reserved inodes. Reserving more inodes after file system is
> > -created requires full file system scan so it can take a long time.
> > +The number must be at least 10, default is 64. Currently 10 is enough but
> > +future features may require additional reserved inodes. Reserving more inodes
> > +after file system is created requires full file system scan so it can take a
> > +long time.
> > .TP
> > .I undo_dir
> > This relation specifies the directory where the undo file should be
> > -- 
> > 2.1.4
> > 
> > --
> > To unsubscribe from this list: send the line "unsubscribe linux-ext4" in
> > the body of a message to majordomo@vger.kernel.org
> > More majordomo info at  http://vger.kernel.org/majordomo-info.html
> 
> 
> Cheers, Andreas
> 
> 
> 
> 
> 
-- 
Jan Kara <jack@suse.com>
SUSE Labs, CR